68086 RESEARCH REPORT 2012 Research at Work Assessing the Influence of World Bank Research Development Economics (DEC) April 11, 2012   TABLE OF CONTENTS Acknowledgments ........................................................................................................................ iii Abbreviations and Acronyms ..................................................................................................... iv Executive Summary ...................................................................................................................... 1 Chapter 1. The World Bank’s Research Publications: Assessing Their Influence on Development Thinking ............................................................................................................... 13 Citation Data and Analysis ..................................................................................................................... 14 World Bank Research Publications and Their Influence ........................................................................ 15 Comparisons with Other Institutions ...................................................................................................... 21 Conclusions ............................................................................................................................................. 26 Chapter 2. Perceptions of Research in World Bank Operations............................................ 27 The Survey .............................................................................................................................................. 28 Familiarity with World Bank Research among the Bank’s Operational Staff ........................................ 28 The Value of Research in Operational Work .......................................................................................... 31 Accessing Research within the Bank ...................................................................................................... 35 Future Demand for Research .................................................................................................................. 39 Role of DEC as Producer and Diffuser of Research ............................................................................... 43 Conclusions ............................................................................................................................................. 46 Chapter 3. Research-Driven Data and Software ..................................................................... 47 Compilation, Collection, and Construction............................................................................................. 48 Conclusions ............................................................................................................................................. 61 ANNEX A. RESEARCH PROGRAMS .................................................................................... 63 A.1 Agriculture and Rural Development Research Program............................................................. 63 A.2 Environment, Energy, and Sustainable Development Research Program .................................. 69 A.3 Finance and Private Sector Development Research Program ..................................................... 76 A.4 Human Development and Public Services Research Program.................................................... 83 A.5 Macroeconomics and Growth Research Program ....................................................................... 91 A.6 Poverty and Inequality Research Program .................................................................................. 99 A.7 Trade and International Integration Research Program ............................................................ 107 ANNEX B. DEC VPU BUDGET ............................................................................................. 117 ii   ACKNOWLEDGMENTS The 2012 Research Report was prepared by Shiva S. Makki with editorial support provided by Lisa Ferraro Parmelee under the supervision of Asli Demirgüç-Kunt. Overall guidance was provided by Justin Yifu Lin. The authors are grateful to Aart Kraay for his active involvement in the drafting of this report, and to Jean-Jacques Dethier, Martin Ravallion, Adam Wagstaff, Shaida Badee, Hans Timmer, Luis Serven, Michael Toman, Peter Lanjouw, Will Martin, and Aaditya Mattoo for their valuable comments; some of them provided feedback multiple times. We are also thankful to Julia Baca, Roula Yazigi, Mathew Mulligan, Nancy Lim, and Grace Sorensen for their help and administrative support. Finally, some very useful comments received from network and regional VPUs have greatly improved this report. Chapter 1 is based on the paper, “The World Bank’s Publication Record,� by Martin Ravallion and Adam Wagstaff, World Bank Policy Research Working Paper Series, no. 5374, 2010. Chapter 2 is based on the paper, “Knowledgeable Bankers? The Demand for Research in World Bank Operations,� by Martin Ravallion, World Bank Policy Research Paper Series, no. WPS 5892, 2011. The survey in Chapter 2 was carried out by Jean- Jacques Dethier and Sharon Felzer. Chapter 3 was drafted by Adam Wagstaff, with contributions from DEC staff. Annex A on Research Programs was compiled by Hedy Sladovich, and individual research program summaries were written by DECRG research managers with contributions from regional and network chief economists and directors. Annex B on the DEC VPU budget is drafted with inputs from Eileen Kirby, Trinidad Angeles, and Jimmy Olazo. iii   ABBREVIATIONS AND ACRONYMS AAA Analytical and Advisory Assistance ARD Agricultural and Rural Development AFR African Region AfDB African Development Bank AsDB Asian Development Bank CAPI Computer-Assisted Personal Interview System CCT Conditional Cash Transfer Program CDD Community-Driven Development CDM Clean Development Mechanism DDA Doha Development Agenda DEC Development Economics DECPI Poverty and Inequality DECRG Development Research Group DPI Database of Political Institutions EAP East Asia and Pacific Department EBRD European Bank for Reconstruction and Development ECA Eastern Europe and Central Asia EM Energy and Mining EPA Export Promotion Agencies ESW Economic and Sector Work FAO Food and Agriculture Organization FDI Foreign Direct Investment FDS Database of Financial Development and Structure FFS Fee for Service FPD Finance and Private Sector FTA Free Trade Areas GAD Global Antidumping Database GS Google Scholar GEP Global Economics Prospects HDN Human Development HNP Health Nutrition and Population IADB Inter-American Development Bank IBRD International Bank for Reconstruction and Development IMF International Monetary Fund ICP International Comparison Program ISA Integrated Surveys on Agriculture LPI Logistics Performance Index LGAF Land Governance Assessment Framework LSE London School of Economics LSMS Living Standards Measurements Study MNA Middle East and North Africa Regional Office MIT Massachusetts Institute of Technology NBER National Bureau of Economic Research NYU New York University OECD Organization for Economic Co-Operation and Development OPCS Operational Policy and Country Services PBP Patient-Based Payments PES Payment for Environmental Services iv PPP Power Purchasing Parity POV Poverty PRM Poverty Reduction & Economic Management PRIME Program Initiatives for Monga Eradication PRWP World Bank Policy Research Working Paper PSG Public Sector Governance RCT Randomized Control Trial RePEc Research Papers in Economics SAR South Asia Regional Office SDN Sustainable Development SDV Social Development SSCI Social Science Citation Index SSRN Social Science Research Network SCOPUS Elsevier-Owned Database on Publications and Citations TTL Task Team Leader TRN Transport UCL University College London UNDP United Nations Development Program URB Urban Development WBI World Bank Institute WDR World Development Report WGI Worldwide Governance Indicators WTO World Trade Organization v RESEARCH REPORT 2012 Research at Work: Assessing the Influence of World Bank Research EXECUTIVE SUMMARY 1. This report on the World Bank research program for FY 2009–2011 focuses on the impact of research on development thinking, how the research is perceived and used within and outside the World Bank, and the contributions of research to the development and improvement of data and analytical tools. This biennial report is prepared for the senior management and the Board of Executive Directors to inform them of the various ways in which Bank research influences development thinking. Box 1 provides a brief summary of the four main objectives of the Bank’s research program and the previous biennial reports on research. 2. World Bank research is carried out under the general direction of the Chief Economist and Senior Vice President of Development Economics (DEC). The Development Research Group (DECRG) is the main in-house unit supplying relevant research findings to Bank operations as well as to external clients. Though not a large department, it produces a substantial share of the Bank’s research and has a high profile internationally in development economics research. DEC also includes the Development Data Group (DECDG) and the Development Prospects Group (DECPG). DECDG leads the Bank’s Open Data Initiative, providing free and easy access to over 7,000 development indicators drawn from high-quality national and international sources. The group works extensively with clients and operational teams to improve the capacity of member countries to produce, use, and disseminate statistical information and data and maintains key data management and retrieval systems for a wide range of datasets, including an extensive library of micro-level data drawn from a variety of surveys worldwide. DECPG provides information and analysis on global trends in the world economy, especially on trade, financial flows, commodity prices, and remittance flows, and the impact of these trends on developing countries. The focus of this report is on Bank-wide research, to which DECRG is a significant—but by no means the only—contributor. 3. The World Bank’s recent reform efforts have focused in large part on redefining its role as a “knowledge bank.� Its research program is at the center of efforts to produce new knowledge on poverty, rural development, finance, macroeconomics and growth, infrastructure, environment, trade, and governance—crucial issues for the economic development of client countries. The program generates articles, books, flagship publication series, data, and analytical tools that form the core of Bank-produced knowledge in development economics. 1 BOX 1. WHAT IS WORLD BANK RESEARCH? World Bank research is designed to generate results with wide applicability across countries or sectors, generally focusing on issues of broader concern than the immediate needs of a particular country, sector, or lending operation. The Bank research program has four specific objectives: 1. To generate knowledge that is primarily a global public good serving the development community. 2. To generate knowledge to guide the Bank’s corporate strategies, policy advice, lending operations,and technical assistance. 3. To generate knowledge to address the specific needs of Bank operations, including assessment of development progress in member countries. 4. To assist in developing research capacity in the Bank’s member countries. Each biennial report to the Board reviews Bank-wide research and evaluates a particular aspect of work toward those objectives: • The FY 2000–2001 report examined research capacity building. An extensive survey of Bank-supported institutions indicated that the Bank’s catalytic role in flexible partnerships provided mechanisms well directed to the needs of beneficiary countries and regions. • The 2002–2003 report examined the impact of research on Bank operations and showed that Bank research is used extensively in economic and sector work (ESW), and that the quality of ESW improves substantially with its use. • The 2004–2005 report focused on an external independent evaluation of Bank research, which concluded that it is crucial for improving the quality of the Bank’s development work, especially if the institution claims to be a knowledge bank. • The 2006–2008 report presented the strategic vision for Bank research in the context of the Bank’s knowledge strategy, as well as a historical overview of the evolving development process. 2 4. By generating new knowledge, the Bank aims not just to reflect development thinking, but to shape it. Records show that the Bank’s research is both abundant and prominent in the field of development. Since 1995, Bank researchers have published over 15,000 scholarly books and articles, many of which have influenced development thinking, as indicated by the number of times these publications have subsequently been cited. The evidence is clear that these publications are used by many clients, including Bank management and operations staff members, policymakers, the donor community, and institutions of education and higher learning. The Bank has also led an expansion of the volume, quality, and accessibility of development data, strengthening the foundation for development knowledge and well-informed policymaking. 5. Putting this body of research to work first and foremost requires that the research itself be influential and of high quality. The first chapter of this report uses bibliometric data—publication counts and citation analysis—to track the volume of World Bank research and how often it is cited, providing evidence on its quality and influence in the broader research and policy community outside the World Bank. Application of research findings in the practical work of the Bank also requires that operational staff be well informed about research relevant to their day-to- day tasks. The second chapter of this report uses a specially commissioned survey to assess the familiarity of senior operational staff with World Bank research, their incentives to draw on research, and the obstacles they face in doing so. Generating new research and putting the findings to work often also require new data and sophisticated analytical tools. But the quality of those data and tools depends on strong supporting research. The third chapter of this report provides a broad discussion of the role of research in improving the quality of data and analytical tools. Two annexes provide an overview of thematic research programs, as well as resources deployed to support research in the DEC VPU. 6. It is important, however, to note here that this report is not comprehensive in its coverage of all activities and products of DEC. Although a few datasets and analytical tools produced and maintained by DECDG and DECPG are discussed, the report is primarily about channels of the impact of research on development thinking in general, and on Bank operational work in particular. The report is also limited in its focus on research outputs in the form of formal publications (Chapter 1) and research-driven data and tools (Chapter 3). As such it does not cover the wide range of other knowledge products developed and disseminated throughout the Bank. Bank Research Publications and Development Thinking 7. World Bank researchers are both prolific and prominent in the field of development. The Bank’s influence as a highly regarded knowledge institution and source of ideas in the field is reflected in a number of important measures, the most significant being the number of its publications that have appeared in scholarly journals. When a research product has been accepted by a peer-reviewed publication covering development, the implication is that it has been considered to be not only of a high standard, but also of potential interest to a large number of people working and writing in the field. Publication of a paper in a scholarly journal thus 3 provides a valid indicator of research quality, while the number of times the paper has been cited elsewhere provides a valid indicator of its impact and influence. The volume of World Bank research publications is large and growing 8. A recent analysis shows that the World Bank has produced over 19,000 publications since 1973. Nearly 2,000 books and 9,000 journal articles have been published by Bank staff, and close to 4,000 working papers have been produced, most of them in the Policy Research Working Papers series. A database of publications compiled from several major bibliographical indices also includes nearly 5,000 book chapters authored by Bank staff and 500 edited volumes published by the Bank. In FY 2009–2011 alone, Bank staff and consultants produced over 3,000 publications, including over 200 books, 1,300 scholarly articles in peer-reviewed journals, and 1,000 Policy Research Working Papers. They also wrote chapters in books and produced Bank flagship series. The recent output evidences a substantial and sustained rise in the number of publications by the Bank that dates back to 1990, with most of the growth coming from journal articles and working papers. 9. This range of research products is aimed at a diverse client base that includes operational staff, policymakers in developing countries, and the development community. While books and flagship reports are popular among policymakers and development practitioners, journal articles and working papers are widely used by academics and development thinkers. World Bank research publications are of high quality and influential 10. World Bank research publications are widely cited in peer-reviewed journals, books, working papers, technical reports, and even dissertations. The more common types of Bank publications—journal articles and working papers—are also those that appear to have had the greatest influence on thinking among economics and development professionals. For instance, in the Research Papers in Economics (RePEc) ranking of research institutions in the field of development, “World Bank� is invariably among the top three (the rest of the top ten include the International Monetary Fund and economics departments of select universities such as Berkeley, Chicago, Harvard, Massachusetts Institute of Technology, Princeton, the London School of Economics, Oxford, and Yale). The Bank ranks first in terms of the number of articles it publishes, as well as in terms of citations in the top development journals. On average, World Bank papers are cited 12 times each, compared to 9 or fewer for the publications of major universities and international agencies. According to RePEc citation analysis, 27 current Bank staff are among the top 10 percent of all authors mapped to the field “development.� RePEc data also show that the Bank’s research department has 7 current and 10 former staff among the top 100 most cited authors in development. 11. Bank staff also authored two of the ten most cited articles in, respectively, the American Economic Review, the Quarterly Journal of Economics, and the Journal of Human Resources; three of the ten most cited articles in the Journal of Development Economics; four of the ten most cited articles in Economic Development and Cultural Change; and the most cited article in 4 Demography. WDRs including, for example, one outlining a clear strategy for fighting poverty and another that proposed a different way of thinking about service delivery have influenced policymakers and development professionals throughout the world, including Bank operations staff. 12. Yet another measure of the influence and impact of World Bank publications comes from download statistics. RePEc data show that the World Bank’s working paper series is the second most downloaded series (1.4 million) via RePEc, behind only the prestigious National Bureau of Economic Research (NBER) working paper series (3.7 million). This is not just because both are large series. On average, the World Bank series actually has slightly more downloads per paper via RePEc than the NBER series (237 versus 207). World Bank research is more focused than that of other institutions on development issues and developing countries 13. The World Bank’s research program is guided by themes and topics that are crucial for economic development. Large, multi-year projects analyze issues to provide strategic direction on critical problems facing Bank operations as well as its clients. Each centers on a theme, giving coherence to the research and its outputs. 14. The World Bank is ahead of most top universities and all other development agencies in the volume of papers it produces on these topics, most notably in economics. In a comparison with the research output of 14 select universities in the United States and the United Kingdom, Bank articles that explicitly mention the name of a developing country outnumber the average by around one and a half times, and they are far in excess of those by other development agencies. In agricultural economics, the Bank ranks second, just behind the Department of Economics at MIT, and it also ranks second in the field of transition economies, just behind the Department of Economics at the University of Chicago. For research on international trade, the Bank again gets the number two slot, just behind MIT. In the field of banking it comes in third, behind IMF and the Federal Reserve Bank of New York. 15. The Bank’s economics research has tended to be more skewed toward topics relevant to developing and transition countries than that of the top five U.S. universities. Among articles on countries not including the United States in the top 200 economics journals, half of the Bank’s publications reported research findings on the poorest 40 percent of countries (ranked by GDP per capita), while those countries accounted for only about one-quarter of the publications of the top five universities. Perceptions of Research in World Bank Operations 16. As background for this report, a specially commissioned survey of the Bank’s senior operational staff aimed to understand how Bank research is perceived by examining their familiarity with it, the value they place on it, and their incentive to learn more about it. While 5 some results suggest room for improvement, a closer look reveals interesting findings regarding the value and use of research by staff distributed across regions, networks, and sectors, as well as many positive points. Research Bank-wide can benefit from this survey and its future updates, not only in improving its products, but also in positioning itself to serve operations better. Valuation and use of research vary across the Bank's operational units. 17. While views vary across the Bank’s operational units, in all VPUs a majority of staff values Bank research. Staff working on poverty, human development, and economic policy tend to value and use research more than those in the more traditional sectors of Bank lending— agriculture and rural development, energy and mining, transport, and urban development. The sector that puts the highest average value on Bank research is the poverty sector within PREM; 93 percent of its survey respondents rated in-house research favorably in terms of its value for their work. VPUs with higher shares of economists and PhDs (in any field) tend to value and use Bank research more. The sectors that use Bank research least also tend to rely less on other (academic) research. Stronger incentive to learn translates into greater familiarity with research 18. The evidence suggests that stronger incentives for learning about World Bank research generally translate into higher investment in learning by staff, as reflected in their familiarity with research. Analysis indicates that about half of the gap between the mean familiarity scores for ARD, EM, TRN, and URB (on the one hand) and EP, EDU, HNP, POV, and SP (on the other) is accounted for by the difference in the value attached to research by staff working in these sectors. In this case, familiarity breeds respect: those staff who are more familiar with in- house research place a greater value on it and are more highly motivated to learn about it, which leads in turn to greater familiarity and more incentive to learn. Operational staff expect their demand for research to increase 19. World Bank staff expect their reliance on research to increase with time—perhaps because they are learning how useful it can be. Among regional VPUs, the demand is expected to rise most, on average, for MNA, while the lowest expected increase is for PREM, where the current level of use of bank research is already high. MNA also has the highest mean score for current reliance on research. The sector where demand is expected to increase most is EM, where usage is low. Both VPUs and sectors that currently rely less on Bank research tend to expect larger growth in usage. 6 World Bank staff values in-house research and favors increasing funding to it 20. Most also believe that the research department is successful in its role as the key in-house unit supplying relevant findings to Bank operations, and they support increasing funding to it rather than turning more to external sources. Support is greater among those who rate highly the technical quality and policy relevance of in-house research and those who think it helps with analytical and advisory assistance (AAA). Most supportive of in-house research—and in favor of increased resources being devoted to it—are those who rely more on it and, especially, those who expect their reliance to increase with time. Researchers’ involvement in operational work is associated with improved familiarity with research and its use 21. The survey analysis shows a positive relationship between the level of cross-support provided by the research department to the VPUs and their familiarity with research. DEC research staff provide analytical support to Bank operations, including economic and sector work (ESW), technical assistance, country program support, lending preparation and project supervision, quality assurance, and training activities. These activities account for 30 percent of the working time of research staff. During the last three fiscal years (2009–2011), DEC researchers collectively have provided, on average, 1,300 staff weeks of cross-support per year, with ESW accounting for a third of that support. Regions received about 70 percent of DEC cross-support during this period. AFR received 22 percent, and EAP and SAR received 11 percent each. Among the sectors, economic policy received the largest share of DEC cross- support (23 percent). Other key sectors that received support included social protection (7 percent), FPD (6 percent), ARD (6 percent), and poverty reduction (6 percent). The level of cross-support provided to VPUs with higher shares of PhDs and of economists per capita was also higher. Survey results suggest ways to improve Bank research 22. Although the majority of operational staff values research, the survey suggests scope for improvements. First, the Bank’s research program needs to continue to improve its alignment with operational needs and trends. Second, more work is needed to understand the factors, on both the supply and demand side, underlying the low levels of familiarity with research and the ineffectiveness of dissemination efforts in certain regions and networks. Third, interactions between researchers and operational staff could be strengthened, not only to increase the use of research in operations, but also to continue to improve the relevance of research for Bank operations. 7 Impact of Research on the Quality of Data and Analytical Tools 23. The discipline of policy analysis is increasingly sophisticated and requires new theories, data, and tools to address emerging challenges. Over the years, the World Bank’s research department has been a major producer and disseminator of all three. The Bank research program plays an important role, not only in improving the quality of existing data and tools, but also in generating new ones. In addition to being at the forefront of an “Open Data and Open Knowledge� policy, Bank researchers have advanced the idea of “wholesaling research�—that is, empowering researchers in the developing world to do high-quality, policy-oriented analytical work themselves. 24. The third chapter of this report focuses on the particular role of research in creating and improving the quality of data and analytical tools. Given this emphasis, it does not cover a number of other, highly visible institutional DEC data products, such as the World Development Indicators (WDI), which is the flagship Bank database for development assembled by DECDG from established international sources. It also does not attempt to describe the many data gathering efforts ongoing in the regional and network units of the Bank. Researchers produce new datasets as an integral part of their research 25. Good research provides the theoretical and practical basis for creating and organizing useful datasets. Over the years, World Bank researchers have pioneered a strategy of collecting data from scratch if interesting policy questions cannot be answered using existing data, with collection efforts spanning almost the full range of Bank themes. In addition to enabling the exploration of key policy questions that could not be answered with the datasets available at the time, many of the Bank’s data collection efforts have made it possible to compare countries and track progress over time and to increase monitoring activities substantially. 26. Agreement is widespread that thoughtful data collection, guided by researchers who understand what types of information are most needed to address fundamental policy questions, is one of the Bank’s greatest contributions to the field of development. The Bank has taken giant strides in improving its work in this area over the past few decades. In 1980, the need for a better understanding of the microfoundations of development led to the creation of the Living Standards Measurement Study (LSMS) at the World Bank. Innovations in survey research methods and new findings from LSMS surveys are combined with information from income and expenditure studies to shed light on the consumption possibilities of poor people and suggest new standards for data collection. Much research undertaken by the Bank is based on microdata. 27. A number of datasets have been constructed by piecing together information from different sources and then performing policy-relevant modeling. Examples include the environmental datasets, which begin with information gleaned from maps and satellite images. The Sea-Level Rise Dataset, which aims to provide information for assessment of the 8 consequences of continued sea-level rise for 84 coastal developing countries, was constructed by using Geographic Information System software to overlay spatially disaggregated global data on critical impact elements, such as land, population, agriculture, urban extent, wetlands, and GDP; the database constructors then model the likely effects of sea-level rise. The Intensification of Storm Surges Dataset adopts the same approach, but models the effects of increased sea surface temperature through intensified cyclone activity and heightened storm surges. 28. The need to measure production and consumption across countries at a common price level led to the International Comparison Program (ICP) through the joint efforts of the University of Pennsylvania, the Ford Foundation, and the World Bank. The World Bank now hosts the global office of the ICP and continues to draw upon the knowledge of researchers inside and outside the Bank to guide the program’s development. The ICP data are widely used to set international poverty lines with common purchasing power, compare relative sizes of economies, and allocate quotas and funds across countries and regions. 29. The World Bank’s highly visible global poverty monitoring efforts—including the well- known “dollar-a-day� measures—have been based in the research department and currently draw on 850 household surveys for 130 developing countries. The Bank’s leadership in this area is widely recognized. For example, The Economist magazine stated in its issue of March 3–9, 2012, that “the best estimates for global poverty come from the World Bank’s Development Research Group,� and proceeded to describe in detail the latest estimates. This work draws on data sources and expertise throughout the World Bank, from other regional banks (through the International Comparison Program), and from developing countries directly. The Bank research program produces new analytical and computational tools 30. Research plays a valuable role in expanding and improving the toolkit routinely employed by policymakers and analysts engaged in the World Bank’s country analytical work and by those in client countries, empowering them to do high-quality analytical work themselves. In response to the Open Knowledge Initiative, the research team expanded its involvement in the development of software products for applied economic research. A newly created computational tools team in DECRG has brought together professionals with expertise spanning high-level programming, advanced econometrics, and geographic information systems. 31. The World Bank has devoted substantial resources—especially in the past five years—to developing software geared to helping researchers and analysts around the world take advantage of the data and methods available to them. ADePT, a software platform for automated economic analysis, is an excellent example of a computational tool invented by the bank to make existing analytical methods of inequality and poverty measurement easy to implement. A multi-module program, ADePT facilitates access to highly sophisticated quantitative methods, many developed by Bank researchers who are global leaders in their fields. PovcalNet allows users to conduct independent poverty analysis by applying their own poverty lines to the same household surveys used in the World Bank’s poverty calculations and using the same computational methods, while also ensuring the Bank honors its agreements with national governments not to distribute the raw data. PovMap, which provides both an innovative analytical tool and computational tools to 9 facilitate its use at all stages of poverty mapping activities, has been implemented in more than 70 countries. 32. The research and development of iSimulate, an open web platform for performing collaborative economic simulations, has allowed researchers to strengthen their forecasting models and their ability to collaborate with regions, easily tapping into the expertise, monitoring, and econometric works of countries around the world. Research Agenda Going Forward 33. The research agenda evolves continually, in response to operational and policy needs. DEC management has identified four major sets of challenges for future research: (i) understanding how access to economic opportunities can be broadened to ensure rapid poverty reduction and greater human development; (ii) improving the means to meet existing and new risks facing economies and people at local, regional, and global levels; (iii) understanding the roles of states, markets, and the private sector in promoting the transformation of economies; and (iv) assessing the effectiveness and results of development efforts, including external assistance. Each of these themes is summarized in Box 2. Bank Investments in Research 34. The Development Economics (DEC) VPU budget allocation remained essentially flat in real terms at around $52 million for the period FY 2009–2011. DEC’s share of the total World Bank net administrative budget during the three fiscal years remained steady at 2.9 percent, but is expected to decline slightly in FY 2012 to 2.8 percent. The allocated budget for the Research Group alone accounts for 1.2 percent of the total Bank net administrative budget (please see Annex B for details). The Bank is under-investing in research and data compared to some other organizations 35. Between FY 2007 and 2012, the DEC VPU budget declined by 5 percent in real terms. This decline in allocated Bank budget has required significant tradeoffs with direct implications for research. Budget pressures have necessitated greater selectivity among research priorities, contributing to some deterioration over recent years of research capacity in some core areas, which can also be gleaned from survey responses. A key challenge for research managers in staffing their teams is to attempt to anticipate the future research needs of clients, while continuing to deliver on what are often multi-year research activities, amid demands for more flexible staffing in the face of budget constraints. Greater emphasis on producing data and tools has also put significant pressure on the budget, since these wholesale research activities are not only resource intensive themselves, but require strong in-house research as an integral part of their production. 10 36. The Bank’s allocation to research and data activities has been rather low compared to that of other large institutions. For example, the IMF’s combined research and data budget is 6.8 percent of its administrative budget, and the United States Federal Reserve Board of Governors spends about 12 percent of its operations budget on research and statistics. 37. While valid comparators beyond similar institutions are difficult to identify, it is interesting to note that the Bank also spends much less on research than the private sector on average spends on research and development activities. For example, OECD data show that in 2008 (the latest year available), manufacturing firms in the United States spent on average 10.5 percent of value added on R&D, while in knowledge-intensive industries such as information technology the share has been as high as 28 percent. Another comparison comes from U.S. federal funding of academic research, which grew by 2.8 percent per year in real terms during the 2000s, and even increased by 22 percent cumulatively between 2006 and 2009, despite tough economic times. By these standards, the Bank is substantially under-investing in research and data creation. 11 BOX 2. RESEARCH AGENDA GOING FORWARD 1. Opportunities • Why do some countries and places attain faster poverty reduction and more inclusive development than others? • Governance challenges in assuring better education, health, and social protection. Local governance and information to strengthen provider incentives. • Focus on gender. • More inclusive global integration and/or regional integration, linked to global markets. • Policies to enhance access to finance. • Rural development and natural resource management. • Clean energy: better data on costs of alternative energy resources. New lessons from policy impact evaluation, research on political economy issues. • Better spatial data on water availability and its economic value in alternative uses. Similar need for improved information on non-market values of ecosystem functions. • Environmentally sustainable longer-term growth—synergies and tradeoffs. . Risks • Understanding the global financial crisis and recovery. • Fragile and conflict-ridden states. • Post-crisis perspectives on macroeconomic management and job creation. • Food price volatility: causes and impacts. • More effective and cost-efficient social protection. • Managing environmental and natural resource scarcity risks. • Dynamics of poverty; everyday “micro risks.� • Mapping hazards: improved spatial data on risks and links to economic development. . Transformations • Links between structural change and broader development goals, including poverty reduction and job creation. • Role of states, markets, and the private sector in promoting economic transformation. • Appropriate policies at each stage of development. • Governance issues for industrial upgrading and structural change. • Role of agriculture, priorities, and tradeoffs. Green revolution in Africa. • Macro data, especially on sectoral composition of output. • Better data on infrastructure investments and their economic impacts. • Firm-level data, for formal and informal firms, especially to study productivity growth and firm– industry dynamics. • Household-level data (archival + new) to link to welfare outcomes. . Results • Monitoring impacts of crisis, including on poverty and human development. • Measures of country performance, including benchmarking and identifying comparators. • Broader approach to “evaluation,� emphasizing external validity, drawing on richer economic modeling, more diverse types of data, and multiple disciplines, tailored to strategic knowledge gaps, pragmatic methods in the above areas—particularly for “large� interventions with economy-wide effects, difficulty in constructing counter-factuals. • Measures of country performance, including benchmarking and identifying comparators. 12 CHAPTER 1. THE WORLD BANK’S RESEARCH PUBLICATIONS: ASSESSING THEIR INFLUENCE ON DEVELOPMENT THINKING1 1.1 The World Bank is one of the most important centers of research in development economics today. Its research program generates books, book chapters, journal articles, reports, working papers, data, and analytical tools that form the core of Bank-produced knowledge in development economics. Its products are widely used internally for economic and sector work, policy dialogue, and technical assistance, and externally for policy advice and education. 1.2 This chapter presents extensive bibliometric data to provide a quantitative and comparative perspective on the Bank’s influence on development thinking. It analyzes not just the Bank’s own publication series, but also articles by Bank staff published in independent peer- reviewed journals, and books and book chapters by Bank staff produced by external publishers— a broader view of Bank publications than is typically adopted within the institution. And, of course, the Bank’s knowledge portfolio is larger than just formal publications. Most of its economic and sector work (ESW) is never published formally, and its “ dissemination� effort, which might take the form of briefing notes or searchable databases, rarely is. When a piece of work does appear in a publication, the implication is that it has been considered—whether by the Bank itself or by a commercial publisher—to be not only of a high standard, but also of potential interest to a large number of people working and writing on development. 1.3 By connecting existing knowledge to policymakers, their advisors, and voters, these publications have a considerable impact on policy and practice. The World Development Report (WDR), familiar to almost everyone working on development, is a good example. As important as its role in disseminating existing knowledge is, however, the Bank is, first and foremost, a generator of new knowledge. It aims not just to reflect development thinking, but to shape it. The staff of its dedicated research department and a large number of staff outside it publish in scholarly journals. For developing countries especially, this may represent a dramatic increase in the knowledge base. Bank authors also create knowledge by developing new ideas and methodological tools and presenting relevant empirical evidence to support them. In short, while “influence on development thinking� is not the only yardstick against which to assess the impact of Bank publications, it is a legitimate and important one. 1.4 In this analysis, two main metrics are used as indicators of the presence and influence of World Bank research in the field of development: counts of the Bank’s publications in the professional literature and counts of the frequency with which these publications are cited by others. It is assumed that a publication considered by others to contribute to new knowledge will get cited, following normal scholarly practices, and that the more knowledge it adds (or the greater the importance of that knowledge), the more often it is likely to get cited. These counts, and comparisons with the publication and citation counts of 14 top universities and 7 other 1 This chapter is based on the paper “The World Bank’s Publication Record,� by Martin Ravallion and Adam Wagstaff, World Bank Policy Research Working Paper Series, no. 5374, 2010. 13 international agencies working on finance and/or development, provide evidence of the widespread influence of Bank research and reflect its contribution to new knowledge in a wide range of areas. This is especially true of publications that focus on development, on economics within specific geographical areas such as Africa, China, and India, and on specific topics such as education, growth, trade, inequality, health, poverty, and finance and the private sector. Citation Data and Analysis 1.5 To measure the volume of publications and the frequency with which they are cited, a dataset of publications and citations to them was assembled, drawing on several major publication and citation databases (see Box 3 for more details on different sources). The data are analyzed to compare different types of World Bank publications with one another, and to compare World Bank publications with those of other institutions. 1.6 Two main measures of the production and influence of World Bank publications are used. Production is measured simply as the publication count, while influence is measured using data on citations. Specifically, the mean and median number of citations per publication are scale-independent measures of average influence; a small or large institution could get a high average citation score. A more sophisticated measure of aggregate scholarly influence through citations is Hirsch’s h-index, which aims to capture both the volume of work and its influence. 2 This index is easy to calculate and interpret. An h-index of x means that the individual (or institution or journal or book series) has published x items, each of which has been cited at least x times (Figure 1.1). Thus, one cannot get a high h-index simply by publishing lots of papers; they also have to have been influential, as indicated by the citations by other authors. Naturally, the h-index of a researcher tends to rise with years of publishing. Scale differences clearly also matter. Other things being equal, the more researchers an institution has, the higher its h-index will be. Average citations and the h-index should, therefore, be looked at alongside one another. 2 J. E. Hirsch, “An Index to Quantify an Individual’s Scientific Research Output,� Proceedings of the National Academy of Sciences of the United States of America 102 (46): 16569, 2005. 14 BOX 3. SOURCES OF PUBLICATIONS AND CITATIONS DATA Data came from the following sources, all of which were searched for publications in which any of the authors listed had indicated the World Bank as their institutional affiliation: • Social Science Citation Index (SSCI). SSCI dates back to 1982 and contains mostly journal articles. • The Elsevier-owned SCOPUS database. Commencing in 1965, SCOPUS covers a longer time period and a larger set of journals than SSCI and includes several non-journal publications; among these are North-Holland Handbook chapters. • EconLit. Dating back to 1969, EconLit contains journal articles but also some book chapters and some working papers. • The World Bank’s e-Library. The e-Library contains Bank-published books, dating back to 1978, and working papers in the Bank’s Policy Research Working Paper (PRWP) series, dating back to 1994. • The U.S. Library of Congress catalogue. The Library of Congress catalogue was searched for books published by the Bank but missing from the e-Library. These were typically jointly published with a university press and presumably omitted from the e-Library for copyright reasons. The Library of Congress books date back to 1974. • Google Scholar (GS). Google Scholar is an internet-based search engine designed to locate scholarly literature available across the Web. Citation data from Google Scholar were included to extend the range of publications covered to articles in new journals, books, working papers, technical reports, and dissertations. Note: Duplicate titles for each publication type were eliminated from the databases constructed from these sources, but not, initially, duplicates across publication types, so a title could appear, for example, as both a working paper and a journal article, and sometimes as a book chapter as well. World Bank Research Publications and Their Influence 1.7 The World Bank’s research program produces a wide range of outputs—books, book chapters, journal articles, reports, working papers, data, and analytical tools—that are widely used internally for economic and sector work, policy dialogue, and technical assistance and externally for policymaking and education. Although a wide range of research products is critical to reach diverse clients, this paper will focus only on books, book chapters, journal articles, and working papers. 1.8 The assessment begins by looking at the Bank’s publications portfolio in terms of size and publication types and presenting data on trends in volume and the differing degrees of influence on development thinking of Bank publications, as indicated by the citation data. 15 Volume of World Bank research publications is large and growing 1.9 The World Bank staff has produced nearly 20,000 publications since the early 1970s, including books, book chapters, journal articles, and working papers. Table 1.1 shows the breakdown of those publications by type. Nearly 9,000 journal articles have been published by Bank staff, over four times the number of Bank-published books. Nearly 4,000 working papers have been produced, most of them in the PRWP series. The database also includes nearly 5,000 book chapters authored by Bank staff and 500 edited volumes published by the Bank; of course, not all of the former appear in the latter. The bulk of the journal articles come from the SCOPUS database; the books and edited books, by contrast, come mostly from the e-Library, as do working papers. Book chapters come largely from EconLit. Figure 1.1. Hirsch’s h-Index—Measuring the Impact of Research Publications 20 18 16 14 One "Home Run" Author Number of Citations 12 Broadly Cited Author 10 Less Cited Author 8 6 H-Index 4 2 0 0 5 10 15 20 Rank of Publication by Citations 16 Table 1.1: Sources and Summary of Bibliographic Metadata on World Bank Publications Conference Edited Working Article Book Chapter Total Proceedings Volume Paper Total 8,834 1,955 4,669 112 506 3,716 19,792 Total with duplicates 8,829 1,900 4,509 102 478 3,003 18,833 eliminated Non-duplicate 6,089 1,465 3,235 59 443 2,992 14,295 publications since 1995 % by research staff 32 10 33 0 2 41 31 Notes: The dataset was assembled starting with SCOPUS, then SSCI, then EconLit, and so on. 1.10 The final row of Table 1.1 shows that nearly 70 percent of Bank publications since 1995 were produced outside the research department; this brings home the fact that staff members outside the research department are also engaged in research. Even so, the research department, with its current staff of under 100 (less than 1 percent of the Bank’s employees worldwide), accounted for 41 percent of working papers, 32 percent of journal articles, 33 percent of book chapters, 10 percent of the Bank’s book catalogue, and 2 percent of edited volumes. 1.11 Figure 1.2 tracks the composition of the Bank’s publications since the early 1970s. The figure shows that the number of books per year has remained fairly constant since the mid-1990s, with some limited growth in edited volumes and book chapters. The big growth has come from journal articles and working papers, particularly since 1990. 17 Figure 1.2: World Bank Publications 1973–2006 (7-year moving averages) 500 Journal Articles 450 Books Chapters 400 Conference Proceedings Edited Volumes 350 Working Papers 300 Publications 250 200 150 100 50 0 1970 1975 1980 1985 1990 1995 2000 2005 2010 Note: The data are presented as seven-year moving averages to depict the trends more clearly. World Bank research publications are of high quality and influential 1.12 World Bank research publications are widely cited in peer-reviewed journals, books, working papers, technical reports, and even dissertations. The more common types of Bank publications—journal articles and working papers—are also those that appear to have had the greatest influence on thinking among academics and development professionals. Table 1.2 reports citations for different types of Bank publications. The percentage cited and the mean and median citations vary by publication type. The types with the highest percentages ever cited and the highest median citations per publication are journal articles and working papers. For World Bank publications, volume and influence on development thinking appear to go hand in hand: the more common types of Bank publications—journal articles and working papers—are also those that appear to have had the greatest influence on thinking among academics and development professionals. This is reflected in the h-indices. World Bank staff have authored 261 journal articles that have been cited at least 261 times, and 114 working papers that have been cited at least 114 times (with some overlap between these categories). These are followed by 133 books that have been cited at least 133 times and 102 book chapters that have been cited at least 102 times. 18 Table 1.2: Citation Counts of World Bank Publications from Google Scholar Book Conference Edited Working Article Book Chapter Proceedings Volume Paper All staff, since 1982 Number of 8,829 1,912 4,509 102 478 3,003 publications % cited 78 68 51 39 70 83 N cites >100 858 175 105 0 18 150 Median citations 10 8 1 0 5 8 Mean citations 42.8 37.9 13.4 4.8 19.2 28.5 h-index 261 133 102 13 50 114 All staff, since 1995 Number of 6,089 1,475 3,235 59 443 2,992 publications % cited 80 66 51 42 69 83 Median citations 11 5 1 0 5 8 Mean citations 45.5 27.5 13.6 7.7 18.5 28.5 h-index 229 95 88 13 46 116 Research staff, since 1995 Number of 1,804 152 1,082 0 10 1,223 publications % cited 86 87 63 -- 70 81 Median citations 21.5 25.5 4 -- 4.5 10 Mean citations 76.6 65.1 24.8 -- 20.1 40.8 h-index 181 53 70 -- 5 98 Sources: Authors’ calculations. Bibliographic data are from the World Bank publications database constructed for this report. Citations for the data are from Google Scholar. 1.13 Table 1.2 also compares the citation statistics of publications authored or edited by research staff with those of non-research staff, from 1995 onward. In general, the fraction of publications with at least one citation is higher among those by research staff, and the difference is quite pronounced for books and book chapters, despite books not being a common form of output for research staff. Publications by research staff also accumulate more citations; for example, the median citation for a book authored by a research staff member is over four times that of one authored by a staff member from another department. These differences suggest that publications from the research department on average contain more new knowledge than those from other Bank units. But it also suggests that some publications from units other than the research department are being cited quite heavily. 1.14 Some Bank publications have been especially highly cited and influential. As already noted, journal articles and working papers have been the Bank’s main channel of influence on thinking among economists and development professionals. Among the Bank’s portfolio of journal articles, Bank staff authored two of the ten most cited articles in, respectively, the American Economic Review, the Quarterly Journal of Economics, and the Journal of Human Resources; three of the ten most cited articles in the Journal of Development Economics; four of the ten most cited articles in Economic Development and Cultural Change; and the most cited article in Demography. 19 1.15 The Bank also produces publications that reach out to practitioners, helping to bridge the gap between cutting-edge research methods and routine applications. Anecdotally, some of these Bank reports—including the WDR—appear to have influenced policymakers and other development stakeholders. A famous instance is the occasion on which Bill Gates was shocked by a chart on mortality in the developing world that he saw in the 1993 WDR on health. In an interview on the U.S. Public Broadcasting Service, Gates said this was his “Aha� moment—the moment that prompted him to set up the Bill and Melinda Gates Foundation and focus its efforts largely on tackling disease in the developing world. Similarly, the 1990 WDR, which outlined a clear strategy for fighting poverty, and the 2004 WDR, which proposed a different way of thinking about service delivery, are often claimed to have influenced policymakers and development professionals, including Bank operations staff. 1.16 Yet another measure of the influence and impact of World Bank publications comes from download statistics. Research Papers in Economics (RePEc) data show that the World Bank’s working paper series is the second most downloaded series (1.4 million) via RePEc, behind only the prestigious National Bureau of Economic Research (NBER) working paper series (3.7 million). This is not just because both are large series, with over 6,000 and 18,000 papers, respectively. On average, the World Bank series actually has slightly more downloads per paper via RePEc than the NBER series (237 versus 207). World Bank research publications are heavily focused on development topics 1.17 World Bank publications naturally focus heavily on development economics and related subjects. Table 1.3 shows the 200 most cited Bank journal articles allocated to various topics. Articles on growth account for 23 percent of the top 200 articles and 35 percent of total citations of these articles. Finance and the private sector account for similar shares of total citations, but the private sector articles are more cited on average. The other topics appear with less frequency in the top 200, and the articles on them are cited less on average. Nonetheless, the Bank’s journal article output includes some highly cited articles in each of the topics listed. 20 Table 1.3: Mean and Total Citations for the Bank’s 200 Most Cited Journal Articles, by Topic Mean Total Share of Topic N Citations Citations Citations Growth 45 1,046 47,075 35 Finance 34 485 16,494 12 Private sector 16 830 13,284 10 Governance 16 621 9,940 7 Trade 16 496 7,934 6 Education 10 782 7,824 6 Inequality 10 642 6,420 5 Agriculture 7 699 4,894 4 Poverty 11 439 4,826 4 Environment 12 389 4,670 3 Public sector 8 538 4,305 3 Health 6 437 2,622 2 Aid 4 593 2,373 2 Labor 2 550 1,099 1 Total 197 679 133,760 Sources: Authors’ calculations. Bibliographic data are from the World Bank publications database constructed for this report. Citations for the data are from Google Scholar. Articles were assigned to a (primary) topic manually on the basis of their titles and, where necessary, abstracts. Comparisons with Other Institutions 1.18 The World Bank’s publication volume and citation data are hard to interpret in isolation. Comparative data show how the Bank fares in comparison with other institutions. These include 14 top universities in the United Kingdom and the United States: Berkeley, Brown, Chicago, Columbia, Cornell, Harvard, the London School of Economics (LSE), the Massachusetts Institute of Technology (MIT), New York University (NYU), Oxford, Princeton, Stanford, University College London (UCL), and Yale; and 7 other international agencies working on finance and/or development: the African Development Bank (AfDB), the Asian Development Bank (AsDB), the European Bank for Reconstruction and Development (EBRD), the Food and Agricultural Organization (FAO), the Inter-American Development Bank (IADB), the International Monetary Fund (IMF), and the United Nations Development Program (UNDP). World Bank research is more focused on development issues and developing countries 1.19 The stronger development focus of the Bank’s journal articles in comparison to other institutions comes through in Figure 1.3, which uses word clouds to display the frequency of the 150 words most commonly used in the titles of articles in the economics, econometrics, and finance group in the SCOPUS database. Calculation of the Herfindahl index (a common measure of concentration) with respect to the SSCI subject areas reveals that the Bank’s research is far more concentrated than that of any of the universities. Even the London School of Economics (which specializes in economics and other social sciences) has a Herfindahl index much lower than the Bank’s. The Bank’s emphasis on agriculture, poverty, inequality, and rural issues is 21 accentuated by the absence of the words “agriculture,� “poverty,� “inequality,� and “rural� from the university word cloud. “Health� and “finance� appear equally prominently in both word clouds; “education,� “trade� and “market� feature in both, but the prior two are more prominent in the Bank’s and the latter in the university word cloud. 1.20 The World Bank is ahead of most of the top universities and all the development agencies in the volume of its papers on development topics, ranking first in growth, poverty, and education; second in health; and third in trade and inequality (Table 1.4). In terms of the number of papers explicitly mentioning the name of a developing country, Berkeley, Harvard, and Oxford all come out ahead of the Bank, but the Bank’s article count is around one and a half times the average among the 14 universities and far in excess of the other development agencies. The Bank ranks third, fifth, and second, respectively, in terms of articles explicitly mentioning Africa, China, and India in their titles. Although Harvard outranks the Bank on all three, and on China the Bank’s count is somewhat less than the average count of the 14 universities, the Bank’s count is more than one and a half times the average count of the universities on Africa and twice the average count on India. 1.21 Bank publications on the poorest countries are especially numerous. Consistent with these findings are those of a 2009 study that used a database of over 76,000 empirical economics papers published since 1985. 3 The study found that, naturally, the Bank’s economics research tended to be more skewed toward low- and middle-income countries than that of the top five U.S. universities (Harvard, MIT, Chicago, Stanford, and Princeton). Focusing on articles on countries not including the United States in the top 200 economics journals brought the numbers of publications between the Bank and these universities closer to one another, but with an important difference: half of the Bank’s publications reported research findings on the poorest 40 percent of countries (ranked by GDP per capita), while those countries accounted for only about one-quarter of the publications of the top five universities. 3 J. Das, Q. T. Do, K. Shaines, and S. Srinivasan, “U.S. and Them: The Geography of Academic Research,� World Bank Policy Research Working Paper Series, no. 5152, 2009. 22 Figure 1.3: Word Clouds for Titles of Journal Articles: World Bank versus Universities a. World Bank b. Universities Source: Articles on “economics, econometrics, and finance� in SCOPUS database. 23 Table 1.4: Journal Articles by Topic “Economics� or “Planning Number of Articles and Development� Topics Institution Health Growth Trade Poverty Inequality Education Africa China India Economics Berkeley 307 439 227 177 81 93 243 368 191 Brown 143 91 84 86 23 27 142 130 77 Chicago 307 263 180 181 83 103 86 197 112 Columbia 432 840 436 327 84 127 347 712 201 Cornell 253 302 273 209 94 105 191 172 104 Harvard 598 692 446 403 161 298 478 524 259 LSE 310 472 184 221 75 91 245 191 160 MIT 265 267 34 88 50 90 30 119 53 NYU 291 296 121 164 51 33 108 121 48 Oxford 332 337 205 198 85 128 539 217 178 Princeton 151 226 78 126 67 59 106 185 65 Stanford 290 398 93 178 84 126 149 321 82 UCL 103 106 125 179 57 35 188 67 72 Yale 188 267 152 110 63 88 186 169 98 World Bank 747 570 475 273 223 187 352 224 248 Rank among universities Article count 1 3 1 3 1 2 3 5 2 h-index 3 10 3 2 2 2 2 3 2 Av. citations 10 11 9 7 8 9 6 3 2 * Includes economics, planning and development, business, finance, and area studies. Source: Authors’ calculations using SSCI data. Quality of Bank research publications in development economics is quite high relative to that of universities and other development organizations 1.22 Table 1.5 takes the sets of SSCI-indexed articles and reports citations of them by other articles included in the SSCI database. The h-index, which captures both the volume of publications and their citations, is calculated for each institution. The index reveals that the top 83 most-cited articles published by World Bank staff are cited at least 83 times. According to this index, the Bank ranks tenth among the universities and first among the international agencies. 1.23 Comparisons within the development field and in the areas in which the Bank specializes are also revealing. Table 1.5 shows citation data for the top 16 journals for development. At first glance, the Bank appears to fare worse in terms of average citations than it does on the number of articles in these 16 journals (position 12, as compared to 1). Its position on the h-index is somewhat better (number 7) but still well below its rank on the volume of papers. The top 16 journals for development include several general journals, however, with limited coverage of development. If attention is restricted to articles in these 16 journals that also fall within the SSCI field of “planning and development,� the Bank’s ranking is much higher: second position on average citations and first position on the h-index. 24 1.24 The last columns of Table 1.5 present SSCI citation data for a list of 27 specialized development journals. Here the Bank ranks first on both average citations and the h-index. On average, World Bank papers are cited 12 times each, compared to 9 or fewer for the publications of major universities and international agencies. Table 1.5: Citation Comparisons across Institutions Using SSCI Data 16 Development Journals 27 Specialized All Journals “Planning and Development All Articles Development� Institution Journals Only Av. Av. Av. h- Av. h- h-index h-index Citations Citations Citations index Citations index Berkeley 18.2 111 33.9 74 12.8 22 7.4 31 Brown 15.3 56 28.4 38 7.7 12 6.8 19 Chicago 33.4 144 54.8 99 3.8 8 4.8 17 Columbia 15.7 86 26.9 55 8.7 10 5.4 22 Cornell 12.0 71 21.0 39 8.5 21 4.9 23 Harvard 26.9 157 46.2 113 9.2 23 8.8 36 LSE 11.0 81 17.3 40 5.3 16 5.6 20 MIT 31.8 136 51.2 102 10.9 17 9.1 24 NYU 17.0 86 33.9 54 2.7 6 5.8 13 Oxford 8.9 79 11.3 41 6.5 24 5.6 29 Princeton 29.1 122 49.3 84 8.3 14 5.9 22 Stanford 23.0 118 41.9 84 8.9 17 8.6 29 UCL 9.4 41 17.2 24 2.5 4 3.3 10 Yale 22.5 99 28.9 56 13.6 19 8.9 28 AfDB 1.1 3 -- -- -- -- 1.0 2 AsDB 4.5 12 -- -- -- -- 7.0 8 EBRD 22.4 11 -- -- -- -- 7.9 4 FAO 2.3 7 -- -- -- -- 2.8 4 IADB 5.0 13 -- -- -- -- 9.4 10 IMF 9.4 52 -- -- -- -- 8.9 36 UNDP 3.0 3 -- -- -- -- 6.5 3 World Bank 13.0 83 18.3 69 13.3 46 12.0 58 Rank among 11 10 12 7 2 1 1 1 universities Rank among international 2 1 -- -- -- -- 1 1 organizations Sources: Authors’ calculations from SSCI data. Bank authors compare well to authors elsewhere 1.25 Research Papers in Economics (RePEc) produces rankings for individual authors. A number of World Bank staff members figure prominently in these rankings, both across all fields and, especially, in the fields of the Bank’s concentration. The Bank currently has 12 authors with h-indices of 12 or higher. This is typical of the universities and other institutions. Although in the 25 Department of Economics at Harvard 19 faculty members have an h-index of 19 or higher, and 25 affiliates of the NBER have h-indices of 25 or higher, all the rest have indices in the 9–13 range. In the field of “development,� the Bank currently has 27 staff members who are among the top 10 percent of all authors. RePEc data also show that the Bank’s research department has 7 current and 10 former staff among the top 100 most cited authors in development. 1.26 In the RePEc ranking of research institutions in the field of development, “World Bank� is consistently among the top three (the rest of the top ten include the International Monetary Fund and economics departments of select universities such as Berkeley, Chicago, Harvard, Massachusetts Institute of Technology, Princeton, the London School of Economics, Oxford, and Yale). The Bank ranks first in terms of the number of articles it publishes, as well as in terms of citations in the top development journals. Conclusions 1.27 In addition to wielding its traditional development aid instruments of lending, policy advice, and technical assistance, the World Bank increasingly emphasizes its role as a generator and disseminator of new knowledge about development—in short, its role as a “knowledge bank.� The main measureable outputs of this role are the Bank’s publications. The Bank has produced a huge volume of books and papers on development—20,000 publications spanning decades, but growing appreciably since 1990. 1.28 World Bank researchers are both prolific and prominent in the field of development. The Bank’s influence as a highly regarded knowledge institution and source of ideas in the field is reflected in a number of important measures, the most significant being the number of its publications that have appeared in scholarly journals. Bank research publications are widely cited in peer-reviewed journals, books, working papers, technical reports, and even dissertations. Of the thousands of World Bank works published since 1973, nearly 70 percent have been cited elsewhere. This indicates widespread influence of Bank research that compares well with other academic and policy research institutions. 1.29 The Bank’s journal articles reflect new knowledge in a wide range of areas. When the focus is on development, on economics within specific geographical areas such as Africa, China, and India, and on specific topics such as education, growth, trade, inequality, health, poverty, and finance and the private sector, the Bank ranks as highly in widely used citation-based metrics as it does in the sheer volume of its published output. Among articles on countries not including the United States in the top 200 economics journals, half of the Bank’s publications report research findings on the poorest 40 percent of countries (ranked by GDP per capita), while those countries account for only about one-quarter of the publications of the top five universities. 1.30 A number of World Bank articles and books have clearly had a considerable impact on development thinking. By generating new knowledge, the Bank aims not just to reflect development thinking, but to shape it. Based on this analysis, the World Bank can legitimately claim to be an innovator in development economics. 26 CHAPTER 2. PERCEPTIONS OF RESEARCH IN WORLD BANK OPERATIONS 4 2.1 Much discussion of the impact of development aid has focused on the role played by the circumstances of recipient countries, notably, whether their policy environments and governance are conducive to high social returns from aid. Less attention has gone to the role played by the staff and managers in international organizations. Concerns specifically about the World Bank have centered on its “lending culture,� which tends to reward operational staff for the volume of their lending, without (it is argued) placing enough weight on its quality. One recent study of aid effectiveness found that the bulk of the variance in the quality of the Bank’s lending operations is within countries rather than among them, suggesting that the effectiveness of the staff in charge of projects on the lender’s side matters at least as much as the policy environment on the recipient’s side. 5 2.2 The effectiveness of operational staff on a given project depends on the knowledge they have available to them for carrying it out. The stock of prior analytical work on a recipient country is a strong predictor of the subsequent quality of lending operations to that country and of development projects there. 6 Whether or not practitioners draw on that stock of knowledge depends on their demand for it as well as its supply. Surprisingly little systematic information has been available about the way Bank research is perceived by the staff in operational units of the Bank. How familiar are they with it? How much do they rely on it for their work? How do the answers to these questions vary across units and sectors of the Bank? Do the Bank’s practitioners value research for their work, and (if so) does this incentive to learn about it translate into greater familiarity with and use of the Bank’s research? 2.3 A survey of the World Bank’s senior operational staff was commissioned for this report, to improve understanding of the way they perceive Bank research by examining both their incentives to use research findings and the responsiveness of the research and researchers to their needs. This chapter discusses in detail survey responses to questions regarding operational staff's familiarity with World Bank research, and the value that they place on it. In addition, the survey asked a number of questions specific to the role of DECRG as a designated research department 4 This chapter is based on the paper “Knowledgeable Bankers? The Demand for Research in World Bank Operations,� by Martin Ravallion, World Bank Policy Research Paper Series, no. WPS 5892, 2011. 5 Denizer Cevdet, Daniel Kaufmann, and Aart Kraay, “Good Countries or Good Projects? Macro and Micro Correlates of World Bank Project Performance,� World Bank Policy Research Working Paper Series, no. 5646, 2011. 6 Klaus Deininger, Lyn Squire, and Swati Basu, “Does Economic Analysis Improve the Quality of Foreign Assistance?� World Bank Economic Review 12 (3): 385–418, 1998; Waly Wane, “The Quality of Foreign Aid: Country Selectivity or Donor Incentives,� World Bank Policy Research Working Paper Series, no. 3325, 2004; Fardoust Shahrokh and Ann Elizabeth Flanagan, “Quality of Knowledge and Quality Financial Assistance: A Quantitative Assessment in the Case of the World Bank,� mimeo, World Bank, 2011. 27 in bringing together internal research capabilities and making it accessible to staff throughout the Bank. Responses to these questions are discussed at the end of this chapter. 2.4 Of course, survey-based perceptions of World Bank research such as those to be described here are not definitive indicators of its actual internal impact or influence. Surveying internal Bank clients is only a first, imperfect step in assessing the impact of the research on operations. 7 Further assessment of both influence and impact will be conducted in future reports. The Survey 2.5 World Bank staff typically work in overlapping teams assigned to specific projects, with each team having a task team leader (TTL). The population for the survey was all senior operations staff, defined as those at grades “GG� and above, who account for about one-quarter of all staff and are all likely to be the TTLs of one or more projects. The survey was administered using a confidential, web-based survey tool. In addition to collecting basic information on vice presidential unit (VPU), sector (essentially synonymous with “network�), and years of Bank service, the survey asked a series of questions on familiarity with World Bank research, sources of knowledge about it, and its value. Of 2,900 recipients of the survey instrument, 555 responded. 8 Familiarity with World Bank Research among the Bank’s Operational Staff 2.6 Learning about research findings is not without cost. To be well-informed, practitioners must spend valuable time finding and studying research products. The costs for doing so will vary across staff, as will the expected benefits. The key question the survey asking about familiarity with and use of World Bank research is, “How familiar would you say you are with World Bank research products/services on a scale from 1 to 10 where 1 means not familiar at all and 10 means extremely familiar?� The mean response is 5.74, with a mode and median of 6. 7 The 2004 Research Report provides the empirical evidence on the importance of research on the quality of Bank’s Economic and Sector Work (ESW). Using data from a large sample of ESW documents, the report shows that Bank research is extensively used in ESW, and more importantly, the analysis shows that the greater the use of research (publication and cross-support), the higher the quality of ESW (as measured by QAG ratings). 8 The background paper (Ravallion, 2011) discusses potential biases stemming from non-random differences in response rates. For example it is possible that survey recipients with strong positive or negative views about research were more likely to register those views by responding to the survey. These biases could work in opposite directions regarding overall perceptions of Bank research. However, as discussed further in that paper, this does not necessarily imply that correlations of responses across questions are biased in one direction or the other. 28 Familiarity with research varies by Bank units 2.7 While familiarity with research is reasonably high for the Bank as a whole, variation across units and sectors of work is substantial (Table 2.1). Among regional VPUs, the highest mean score for familiarity is for the Middle East and North Africa (MNA), while the lowest is for East Asia and the Pacific (EAP). Defining “high familiarity� as above 5, we find that over two-thirds of respondents in MNA rate themselves as having high familiarity, as compared to less than 45 percent in EAP. The differences among network VPUs are even more striking, with only 36 percent of WBI staff rating their familiarity over 5, as compared to 81 percent in HDN and 71 percent in PRM. Table 2.1: Summary Statistics on Familiarity with Research across Vice Presidencies Average Familiarity with Proportion Highly Research Familiar with Research Vice Presidency (10-Point Scale) (% above 5 on Scale) Mean St. error Mean St. error Regional VPUs Sub-Saharan Africa (AFR) 5.67 0.21 52.13 5.21 East Asia and Pacific (EAP) 5.31 0.27 44.62 6.24 Eastern Europe and Central Asia (ECA) 5.41 0.29 47.46 6.58 Latin America and Caribbean (LCR) 5.34 0.24 49.32 5.92 Middle East and North Africa (MNA) 6.13 0.44 66.67 9.74 South Asia (SAR) 5.88 0.30 60.00 7.01 Network VPUs Finance and Private Sector (FPD) 5.50 0.56 50.00 11.92 Human Development (HDN) 7.33 0.45 80.95 8.67 Operational Policy (OPC) 5.41 0.58 47.06 12.25 Poverty Reduction and Econ. Mgt. (PRM) 6.89 0.42 71.43 7.73 Sustainable Development (SDN) 5.63 0.42 50.00 8.00 World Bank Institute (WBI) 5.18 0.88 36.36 14.68 World Bank 5.74 0.10 53.76 2.19 Note: The totals include some other VPUs with insufficient sample sizes. Familiarity with research varies by sector 2.8 The differences among the Bank’s sectors come out sharply in Table 2.2. Over 80 percent of responding staff in the poverty unit (POV) of PRM rate themselves as highly familiar with Bank research, with a mean score of almost 7. Yet less than one-third of those in the energy and mining (EM) and urban development (URB) sectors have high familiarity. A tendency is clear among such traditional “infrastructure sectors�—EM, URB, agriculture and rural development (ARD), and transport (TRN)—to have lower familiarity than do the newer “economic and human development sectors�—POV, economic policy (EP), education (EDU), health and nutrition (HNP), and social protection (SP). The mean score for the traditional sectors is 5.19, as compared to 6.48 for the newer ones—a significant difference. The corresponding proportions of staff with high familiarity scores are 45 percent and 71 percent. In all, 41 staff members in the traditional (“hard� infrastructure) sectors give a score of above 5, while 279 do so in the sample 29 as a whole. Thus, the hard sectors account for 15 percent of the staff with high familiarity. By contrast, these sectors account for 45 percent of Bank lending. Table 2.2: Summary Statistics on Familiarity with Research across Sectors Proportion Highly DECRG Familiarity with Familiar with Staff Research Sector Allocatio (10-Point Scale) Research (% above 5 on Scale) n (%) Mean St. error Mean St. error Agriculture and Rural Development (ARD) 9.01 5.47 0.31 50.00 8.22 Economic Policy (EP) 29.73 6.50 0.19 67.29 4.59 Education (EDU) 3.60 6.46 0.42 68.57 7.95 Energy and Mining (EM) 4.05 4.59 0.52 31.82 10.06 Environment (ENV) 4.95 5.32 0.48 45.45 10.75 Finance and Private Sector (FPD) 11.71 5.70 0.36 50.00 8.01 Health, Nutrition, and Population (HNP) 4.05 5.92 0.36 61.11 8.23 Poverty (POV) 15.32 6.97 0.30 83.33 6.89 Public Sector Governance (PSG) 6.31 5.28 0.44 37.50 8.67 Social Development (SDV) 3.15 5.23 0.31 38.46 13.67 Social Protection (SP) 6.31 6.71 0.57 64.71 11.74 Transport (TRN) 0.45 5.45 0.66 45.00 11.27 Urban Development (URB) 1.35 5.05 0.58 31.58 10.80 Total 100.00 5.74 0.10 53.76 2.19 Note: The totals include some other sectors with insufficient sample sizes. Familiarity with research varies with years of service and location of staff 2.9 Familiarity with World Bank research also varies with years of service, and this effect is markedly nonlinear. Familiarity rises with years of service up to 16 years and falls after that. The median years of service is 10; 20 percent of respondents had more than 16 years of service. Familiarity with World Bank research also differs between those based at the Bank’s headquarters in Washington, DC (where the research department is located), and those in the field offices in developing countries; 65 percent of respondents are in the former group, versus 34 percent in field offices (the remaining 1 percent are in World Bank offices in developed countries, mainly Europe). The fact that familiarity is lower for those in other-country offices suggests that advances in communication technology—for example, the World Bank’s well- developed intranet—have not eliminated the advantages of physical proximity. Uneven familiarity with Bank research across sectors is not easily justified 2.10 One might argue that differences in familiarity with World Bank research just reflect the nature of the work done by operational staff in different sectors. According to this view, staff in the hard infrastructure sectors do not need the Bank’s research as much as those in the social (“soft�) sectors do. To paraphrase one argument that has been made, “We all know that roads are 30 good for development, and we don’t need economic analysis to confirm that; what we need is to make sure the roads meet sound engineering standards.� 2.11 The validity of this view is hard to judge. The nature of the knowledge needed by staff in different sectors clearly differs, but the need for research is surely common among them. Economic analysis is important in formulating strategies and policies for infrastructure, notably in assessing the effectiveness of lending operations (both before and after they have been implemented) and in devising sound pricing and investment policies. And the tradeoffs invariably faced by governments between building roads and building schools (say) must be informed by economic analysis. The uneven familiarity with World Bank research across sectors evident in Table 2.2 is not easily justified from the point of view of sound development policy and Bank lending. 2.12 Some might conjecture that the dominance of economics as a discipline in Bank research makes that research less relevant to some sectors. As discussed above, this does not seem credible as an explanation for why Bank research is used less in the traditional infrastructure sectors. Nor does it fit well with the fact that staff in HNP—dominated by non-economists—are very familiar with research. Possibly the idea applies more to the SDV sector, which uses non- economics social sciences more than economics (note that SDV has the third lowest mean familiarity score in Table 2.2). This division of labor is not, however, imposed from without, but has evolved in a path-dependent way through hiring decisions. And it can change, albeit slowly. For example, DECRG’s latest Policy Research Report (its flagship publication) is a substantial evaluative study of the main lending instruments used by SDV operations. This report stemmed from demands from operations for more knowledge about the impacts of these interventions. The Value of Research in Operational Work 2.13 The incentive to learn about World Bank research is greatest when it is perceived to be of high value to operational staff in their work. The survey asked respondents to rate the overall value of Bank research to their work on a scale of 1 (“not valuable at all�) to 10 (“extremely valuable�). As with other variables, results for the value question show a marked bi-modality, with a high-value mode around 7–8 and a low-value mode around 3 (Figure 2.1). The mean score is 5.61, and the median is 6. Defining high value as a score of 5 or more, we find two-thirds of respondents attach high value to World Bank research. The majority of Bank staff views research favorably 2.14 Tables 2.3 and 2.4 show the differences in this assessment by VPUs and sectors. In all VPUs, a majority give responses of 5 or higher. The VPU giving the highest value is WBI, with an average score of 7 and all respondents rating Bank research highly. The next highest is PRM, with a mean score of 6.5. The sectoral VPUs generally put higher value on research than the regional VPUs. The regional VPU with the highest mean score is (again) MNA. 31 2.15 The traditional “hard infrastructure� sectors tend to put a lower average value on research for their work. The mean score for ARD, EM, TRN, and URB is 4.82, as compared to 6.27 for EP, EDU, HNP, POV, and SP—a significant difference. The sector assigning the lowest score to Bank research in terms of its value to their work is EM (a mean score of 4.1, with only 41 percent rating it highly), followed by TRN, URB, and ARD. The sector giving the highest mean value is the poverty sector within PRM, with a score of 7 and 93 percent of respondents rating Bank research favorably. Figure 2.1: Densities of Perceived Value of WB Research to Respondent’s Work 32 Table 2.3: Summary Statistics on Perceived Value of Research to Work across Vice- Presidencies Value of Research Value Attached to Research Vice Presidency (10-Point Scale) (% Rating 5 or Higher) Mean St. error Mean St. error Regional VPUs Sub-Saharan Africa 5.34 0.25 61.70 5.07 East Asia and Pacific 5.37 0.32 58.46 6.18 Eastern Europe and Central Asia 5.31 0.32 64.41 6.31 Latin America and Caribbean 5.59 0.30 67.12 5.56 Middle East and North Africa 5.92 0.50 70.83 9.39 South Asia 5.74 0.35 64.00 6.87 Network VPUs Finance and Private Sector 6.06 0.75 72.22 10.68 Human Development 5.90 0.57 66.67 10.41 Operational Policy 5.47 0.48 58.82 12.08 Poverty Reduction and Econ. Mgt. 6.46 0.42 77.14 7.18 Sustainable Development 5.00 0.40 55.00 7.96 World Bank Institute 7.00 0.48 100.00 0.00 World Bank 5.61 0.11 65.51 2.09 Note: The totals include some other VPUs with insufficient sample sizes. Table 2.4: Summary Statistics on Perceived Value of Research across Sectors Value Attached to Value of Research (10- Research Sector Point Scale) (% Rating 5 or Higher) Mean St. error Mean St. error Agriculture and Rural Development 5.16 0.42 57.89 8.11 Economic Policy 6.24 0.21 80.37 3.89 Education 5.85 0.49 60.00 8.39 Energy and Mining 4.10 0.51 40.91 10.62 Environment 5.38 0.57 54.55 10.75 Finance and Private Sector 5.49 0.42 62.50 7.75 Health, Nutrition, and Population 6.17 0.40 75.00 7.31 Poverty 7.37 0.37 93.33 4.61 Public Sector Governance 4.90 0.41 53.13 8.93 Social Development 5.46 0.44 69.23 12.96 Social Protection 5.47 0.58 64.71 11.74 Transport 4.88 0.72 60.00 11.09 Urban Development 4.89 0.49 52.63 11.60 Total 5.74 0.11 65.51 2.09 Note: The totals include some other sectors with insufficient sample sizes. Grouping of staff shows relationship between familiarity and perceived value of research 2.16 A cross-tabulation of familiarity with and the perceived value of World Bank research to respondents’ work allows for the grouping of staff into four categories (Table 2.5). The “functionally well-informed� are those who have high familiarity with World Bank research products and feel that Bank research is important for their work. They comprise 42 percent of all 33 respondents and roughly two-thirds of all those with high familiarity. The “independently well- informed,� who are familiar with Bank research but do not view it favorably, are a relatively small group. Among those with low perceived familiarity, slightly less than half put low value on research for their work (the “happily uninformed,� comprising roughly a quarter of all respondents), while the rest—the “frustrated uninformed,� comprising about another quarter of all respondents—think research is valuable, but are not sufficiently familiar with it. Table 2.5: Perceived Value of World Bank Research for the Staff Member’s Work Perceived Value of World Bank Research Familiarity/Value Low Value High Value “Frustrated uninformed� “Happily uninformed� N=123 (23.70%) N=117 (22.54%) Low Mean familiarity=3.35 Mean familiarity=4.00 Mean value=6.35 Familiarity with Mean value=2.72 World Bank research “Independently well-informed� “Functionally well-informed� N=62 (11.95%) N=217 (41.81%) High Mean familiarity=7.26 Mean familiarity=7.58 Mean value=2.90 Mean value=7.57 2.17 The “frustrated uninformed� can be thought of as a group to be specially targeted for research dissemination. Tables 2.6 and 2.7 give the VPU and sectoral distributions of the four types of staff among them. WBI has (by far) the highest share of “frustrated uninformed� staff, with 65 percent in this category, the rest being “functionally well-informed.� The MNA region has the lowest, at 13 percent. By sectors, the transport staff have the highest share of “frustrated uninformed,� at 35 percent, with URB and SDV close behind. EDU has the lowest, at 9 percent. The share of “functionally well-informed� varies from 35 percent in SDN to 63 percent in PRM, while among sectors it varies from 14 percent in energy and mining to 83 percent in the poverty sector. 2.18 A closer look at the “happily uninformed� does not suggest they have simply switched their favor to external (non–World Bank) sources of research. They give a relatively low rating to academia as a source—a mean score of 5.53, significantly different from the 6.87 for all other staff. They also rate consultants significantly lower than average, at a mean of 6.07, as compared to 6.54 for the rest. At the same time, their desire to increase their usage of the Bank’s research is above average; their mean difference between expected usage and current usage (both on the 10- point scale) is 0.80, versus 0.54 for the rest of the staff and 0.43 for the “functionally well- informed.� Evidently, the “happily uninformed� are not entirely happy with the status quo. 34 Table 2.6: Percentage Distribution of Types of Staff across Vice Presidencies “Happily “Frustrated “Independently “Functionally Vice Presidency Uninformed� Uninformed� Well-Informed� Well-Informed� Total Regional VPUs Sub-Saharan Africa 23.40 24.47 14.89 37.23 100.00 East Asia and Pacific 32.31 23.08 9.23 35.38 100.00 Eastern Europe and Central Asia 25.42 27.12 10.17 37.29 100.00 Latin America and Caribbean 21.92 28.77 10.96 38.36 100.00 Middle East and North Africa 20.83 12.50 8.33 58.33 100.00 South Asia 22.00 18.00 14.00 46.00 100.00 Network VPUs Finance and Private Sector 22.22 27.78 5.56 44.44 100.00 Human Development 4.76 14.29 28.57 52.38 100.00 Operational Policy 29.41 23.53 11.76 35.29 100.00 Poverty Reduction and 14.29 14.29 8.57 62.86 100.00 Economic Management Sustainable Development 30.00 20.00 15.00 35.00 100.00 World Bank Institute 0.00 63.64 0.00 36.36 100.00 Total 22.54 23.70 11.95 41.81 100.00 Note: The totals include some other VPUs with insufficient sample sizes. Table 2.7: Distribution of Types of Staff across Sectors “Happily “Frustrated “Independently “Functionally Sectors uninformed� uninformed� well-informed� well-informed� Total Agriculture and Rural Development 34.21 15.79 7.89 42.11 100.00 Economic Policy 10.28 22.43 9.35 57.94 100.00 Education 22.86 8.57 17.14 51.43 100.00 Energy and Mining 40.91 27.27 18.18 13.64 100.00 Environment 36.36 18.18 9.09 36.36 100.00 Finance and Private Sector 25.00 25.00 12.50 37.50 100.00 Health, Nutrition, and Population 11.11 27.78 13.89 47.22 100.00 Poverty 6.67 10.00 0.00 83.33 100.00 Public Sector Governance 37.50 25.00 9.38 28.13 100.00 Social Development 30.77 30.77 0.00 38.46 100.00 Social Protection 11.76 23.53 23.53 41.18 100.00 Transport 20.00 35.00 20.00 25.00 100.00 Urban Development 36.84 31.58 10.53 21.05 100.00 Total 22.54 23.70 11.95 41.81 100.00 Note: The totals include some other sectors with insufficient sample sizes. Accessing Research within the Bank 2.19 How do the Bank’s practitioners obtain the research they use? In addition to papers and reports, they have available to them learning programs, seminars, formal and informal contacts with researchers, and formal and informal networks. The survey data on self-assessed familiarity with Bank research reveal some sources of friction in the diffusion process for research findings, however. 35 Bank staff access research from a variety of sources 2.20 A range of channels for accessing research is addressed by the following question on the survey: “When obtaining research from the World Bank, how frequently do you get the research from…?� Respondents were asked to rate a list of options on a 10-point scale. As Table 2.8 shows, the most highly rated sources of World Bank research are “reports or papers,� the Bank’s intranet, flagship reports, and the main working paper series. “Seminars, workshops, or presentations� and authored reports are close behind. 2.21 Frequency of use for various sources of knowledge need not accord well with their impacts on knowledge. The respondents’ own assessments of their familiarity with Bank research are related to the relative importance of these sources, as well as their own work environments (for example, location, sector, VPU, relationship with researchers, and so forth). For example, when a researcher provides cross-support to a TTL in operations, the TTL will probably get a deeper understanding and appreciation of the contributions of research to operational work than would be gained from reading a final research product that is less distantly related to his or her needs. In other words, the TTL may read more papers, but the cross-support has greater impact. Table 2.8: Sources of Knowledge on World Bank Research Standard Source of Knowledge Mean Median Deviation Informal discussion with researcher(s) 4.50 5 2.72 Hiring researchers (consultants) 4.96 5 2.70 Report(s) or paper(s) 6.68 7 2.27 DECRG researcher cross-support 3.69 3 2.56 Other researcher from network or other part of Bank 5.02 5 2.54 Product available on the intranet 6.52 7 2.44 Seminar, workshop, or presentation 5.85 6 2.36 Authored report(s) 5.87 6 2.55 Flagship report, such as WDR, PRR 6.53 7 2.49 Policy Research Working Paper series 6.37 7 2.45 Journal article authored by Bank staff 4.60 4 2.70 2.22 To capture this complex interactive effect, the study used regression analysis of how the familiarity scores varied depending on sources of knowledge while filtering out the effects of location, sector, and VPU. In this analysis, the use of researchers in cross-support indeed emerges as the most important factor in explaining the differences in overall familiarity with World Bank research. Informal discussions with researchers are also important. The direct (formal and informal) interactions with researchers (including cross-support) have high weights in affecting familiarity with World Bank research, even though they score relatively poorly in the mean ratings for frequency of their use given in Table 2.8. 36 Performance of networks in connecting to research varies across sectors 2.23 Another channel for learning about research is the official networks the Bank has created. The networks are intended to play a key role in linking sources of knowledge (including Bank research) with practitioners within the Bank. To assess how well they serve this function, the survey asked: “To what extent does the network that you belong to help you navigate the World Bank’s body of research and researchers when you request information?� Again, responses are on a scale from 1 (“not at all�) to 10 (“very much�). The mean response is 4.86, and the median is 5. The modal value is 1, with 15 percent of respondents giving this answer. 2.24 The differences across sectors can be seen in Table 2.9. The most highly rated is education; 70 percent of those mapped to the sector rate network performance in assisting with access to research as above 5, as compared to 42 percent of all staff. The next highest mean score is from HNP, with POV coming third; the sector with the lowest mean score is ENV. The positive correlation of the performance ratings of the networks with familiarity with Bank research indicates that having a network that performs poorly adds friction to the diffusion process for research findings. Table 2.9: Staff Assessments of the Performance of Networks in Connecting to Research Average Rating High Rating Sector (10-Point Scale) (% above 5 on Scale) Mean St. error Mean St. error Agriculture and Rural Development 4.59 0.46 24.14 8.08 Economic Policy 4.93 0.28 48.84 5.48 Education 6.33 0.51 70.00 8.50 Energy and Mining 4.78 0.65 38.89 11.68 Environment 4.00 0.66 33.33 12.37 Finance and Private Sector 4.26 0.48 35.29 8.33 Health, Nutrition, and Population 5.63 0.46 53.33 9.26 Poverty 5.48 0.47 48.15 9.77 Public Sector Governance 5.16 0.55 44.00 10.09 Social Development 4.10 0.78 30.00 14.73 Social Protection 5.29 0.73 57.14 13.44 Transport 4.71 0.74 29.41 11.23 Urban Development 5.38 0.60 43.75 12.60 Total: 4.86 0.13 41.71 2.44 Note: The totals include some other sectors with insufficient sample sizes. Diffusion of research varies with internal capacity 2.25 The differences in the extent to which various VPUs draw on the Bank’s research can be expected to reflect differences in their staffing. The share of senior staff who have “economist� in their job titles may matter (Table 2.10), because World Bank research is dominated by economic analysis; and staff with PhDs in any subject area can be expected to be trained in 37 research methods. The greater ability of a VPU with more economists and/or PhDs to conduct its own research may result in low demand for research. On the other hand, a unit’s internal capacity for research may actually enhance its capacity for learning and innovation—an attribute the literature on industrial organization calls its absorptive capacity. Internal capacity reduces search costs and creates the kinds of overlaps in knowledge that facilitate the assimilation of new knowledge from external sources. 2.26 The idea of absorptive capacity is consistent with analysis of the survey data, which indicates that the average familiarity with Bank research of the staff in a VPU is, indeed, significantly correlated with the proportion of staff with PhDs (correlation coefficient of 0.66). On comparing the proportion of PhDs with the high familiarity scores, one finds a correlation coefficient of 0.54, which is also statistically significant. The correlation between average familiarity with research and the share of economists is even stronger. The correlation coefficients between the ratio of economists to senior staff who were sent the survey and the mean score for familiarity is 0.84, while the correlation with high familiarity (score of 5 or above) is 0.75; these are both statistically significant. Economists and staff with PhDs apparently play a key role in the diffusion process for World Bank research throughout the institution. 38 Table 2.10: Economists and PhDs on Staff and Amount of Research Support Annual Cross- Annual Cross- No. Staff with Support by Support Per % Staff “Economist� in Research Capita of Staff Vice Presidency GG+ with Job Title in Department GG+ PhD Staff Directory $’000 ($/Person/ (3-Year Mean) Year) Regional VPUs Sub-Saharan Africa 46.9 57 764 1,174 East Asia and Pacific 44.5 31 460 1,318 Eastern Europe and Central Asia 49.4 43 303 900 Latin America and Caribbean 55.0 42 553 1,790 Middle East and North Africa 48.4 25 281 1,383 South Asia 46.8 22 429 1,296 Network VPUs Finance and Private Sector 38.0 8 189 1,661 Human Development 61.5 38 111 1,437 Operational Policy 49.7 10 32 386 Poverty Reduction and 75.4 56 332 4,368 Economic Management Sustainable Development 42.3 11 374 1,278 World Bank Institute 54.9 11 81 1,084 Development Economics 77.2 110 n.a. n.a. Future Demand for Research 2.27 The demand for research by World Bank operational staff is changing. The survey asked two questions to get an idea of the extent of this change: “To what extent do you currently rely on Bank research for your work?� and “To what extent do you expect to rely on Bank research for your work in the next few years?� Responses are on a scale of 1 (“not at all�) to 10 (“very much�). 39 Operational staff expect their demand for research to increase 2.28 As with other questions, the survey responses on current reliance on Bank research are strikingly bi-modal—low reliance on research and high reliance on research. The average response for the staff who are less reliant on research is around 3, and that of the high-reliance mode around 7. The overall mean response is 5.4 and the median is 6, though this might be considered deceptive given the bi-modality. Defining the low-reliance and high-reliance groups as those with a score of 5 or lower versus above 5 splits the sample almost equally, with 51 percent in the high group. Equally striking is the fact that the lower mode largely vanishes when one turns to the question about future reliance on Bank research; expected reliance is noticeably higher than current reliance. 2.29 Tables 2.11 and 2.12 present a closer look at the sources of this expected increase in demand for Bank research. Among regional VPUs, the demand is expected to rise most, on average, for MNA, while the lowest expected increase is for PRM. MNA also has the highest mean score for current reliance on research. The sector where demand is expected to increase most is EM, and the least increase is expected for SP. Both VPUs and sectors that currently rely less on Bank research tend to expect larger growth in usage. 2.30 This tendency notwithstanding, expected reliance on World Bank research remains significantly lower for the sectors whose current familiarity and use are low. The mean score for future reliance among staff in the seven sectors with lowest mean familiarity scores—ARD, EM, ENV, PSG, SDV, TRN, and URB—is 5.55, as compared to 6.29 for all other staff. The mean expected reliance on research among the traditional “hard� sectors (ARD, EM, TRN, and URB) is 5.34, as compared to 6.72 for EP, EDU, HNP, POV, and SP—again, a significant difference. 40 Table 2.11: Current and Expected Demand for Research across Vice Presidencies Expected Increase in Current Reliance on Demand (Score for Expected Vice Presidency Research (10-point Scale) Demand minus Score for Current) Mean St. error Mean St. error Regional VPUs Sub-Saharan Africa 5.28 0.28 0.55 0.18 East Asia and Pacific 4.98 0.33 0.51 0.15 Eastern Europe and Central Asia 5.45 0.34 0.48 0.14 Latin America and Caribbean 5.18 0.33 0.71 0.18 Middle East and North Africa 6.05 0.48 0.95 0.26 South Asia 5.35 0.39 0.80 0.15 Network VPUs Finance and Private Sector 4.87 0.76 0.43 0.28 Human Development 6.38 0.54 0.43 0.27 Operational Policy 4.46 0.73 0.92 0.64 Poverty Reduction and Econ. Mgt. 6.45 0.45 0.37 0.16 Sustainable Development 5.19 0.43 0.63 0.25 World Bank Institute 5.57 0.79 0.43 0.35 Total 5.40 0.12 0.59 0.06 Table 2.12: Current and Expected Demand for Research across Sectors Expected Increase in Current Reliance on Demand (Score for Expected Sector Research (10-Point Scale) Demand minus Score for Current) Mean St. error Mean St. error Agriculture and Rural Development 5.43 0.55 0.40 0.16 Economic Policy 6.18 0.26 0.72 0.14 Education 5.58 0.46 0.61 0.23 Energy and Mining 3.55 0.38 0.74 0.21 Environment 5.82 0.69 0.35 0.22 Finance and Private Sector 4.94 0.41 0.58 0.14 Health, Nutrition, and Population 5.64 0.41 0.33 0.13 Poverty 6.96 0.34 0.67 0.24 Public Sector Governance 5.31 0.44 0.50 0.33 Social Development 4.91 0.53 0.64 0.33 Social Protection 5.87 0.62 0.20 0.17 Transport 4.67 0.68 0.65 0.27 Urban Development 5.12 0.48 0.29 0.45 Total 5.40 0.12 0.59 0.06 41 Stronger incentive to learn translates into greater familiarity with research 2.31 The evidence suggests that stronger incentives for learning generally translate into higher investment by staff in learning, as reflected in the familiarity with research (see Figure 2.2). Those respondents who believe they have a high familiarity with World Bank research—again defined as a score of above 5—give a mean assessed value of it to their work of 6.53, as compared to 4.49 for those who acknowledge they are not very familiar with the research; the difference is highly significant. Among those familiar with Bank research, 69 percent rate it of high overall value, compared to 35 percent for those relatively unfamiliar. Further analysis indicates that about half of the gap between the mean familiarity scores for ARD, EM, TRN, and URB (on the one hand) and EP, EDU, HNP, POV, and SP (on the other) is accounted for by the difference in the value attached to research by staff working in these sectors. The fact that knowledge responds to incentives for learning suggests that the Bank’s internal system for research and its diffusion is working to create internal absorptive capacity. 2.32 Other evidence on the role played by incentives for learning can be found in how the demand for paid cross-support from the Bank’s research department varies with answers to the value question. Across VPUs, the mean score for the value of World Bank research is highly correlated with the per capita demand for cross-support from DECRG. The correlation coefficient between the three-year mean of cross-support per staff member (grade level GG and above) and the mean value score is 0.72, and it is also 0.72 with the proportion of VPU staff, giving a value score of over 5. This is consistent with the view that incentives for learning in the Bank do generate a response. 42 Figure 2.2: Relationship between Familiarity with and Perceived Value of Research Role of DEC as Producer and Diffuser of Research Bank staff support more resources for in-house research 2.33 Most staff believe that DECRG is successful in its role as the key in-house unit supplying relevant research findings to Bank operations, and they support increasing funding to the department rather than turning more to external sources. The survey asked, “Which of the following statements best represents your views about research at the Bank? (1) I would find it more valuable and useful if the institution were to out-source more of its research; (2) I would find it more valuable and useful if the institution committed more resources to research within an in-house research department.� Overall, 63 percent of respondents support more resources for in-house research. The VPU with the highest support is WBI, at 86 percent; ECA, with the lowest, still has a majority of 57 percent in favor of increased funding. Across sectors, the highest support for in-house research comes from EP and POV, both with 82 percent; the lowest comes from EM, with 40 percent. 2.34 The survey shows that those who rely more on in-house research tend to be more supportive of increased resources being devoted to it, as do those who are more familiar with it. An even stronger correlation is found between support for in-house research and expected 43 reliance on it. Support is also greater among those who rate highly the technical quality and policy relevance of in-house research and those who think it helps with AAA. Finally, those who have a less favorable view of external consultants tend to be more supportive of the research department. Assessments of responsiveness of researchers vary widely across units 2.35 The direct responsiveness of the research department to operational demands is a potentially important channel for knowledge diffusion in the World Bank, but how well does it work? To explore this, the survey asked two questions about the experiences of staff in working with DECRG: “To what extent are researchers at DEC responsive to your needs?� and “To what extent are DEC researchers responsive and available for cross-support?� Again, respondents were asked to rate their experiences on a scale from 1 (“not at all�) to 10 (“to a very high degree�). The first question also provided an option for “depends on the researcher; some are responsive, some are not.� 2.36 Again, the responses are bi-modal, with a larger subgroup of staff centered on an answer of 7–8 and a smaller mass centered on 2–3. The mean and median responses to the first question are 5.8 and 6 respectively; 12.9 percent of respondents gave the answer “depends on the researcher.� The mean and median for the second question are 5.47 and 6, respectively. 2.37 Tables 2.13 and 2.14 give the breakdown of these indicators across VPUs and sectors. The proportion rating DECRG researchers as 5 or higher on their responsiveness varies from 50 percent in OPC to 100 percent in WBI. The proportion with 5 or higher for availability for cross- support varies from 36 percent in South Asia (SAR) to 79 percent in ECA. The differences across sectors in assessments of responsiveness are particularly striking, with 20 percent of staff in SDV rating responsiveness as high, compared to 90 percent in POV and 86 percent in EP. Researchers’ involvement in operational work and the use of research go hand-in-hand 2.38 The use of DECRG staff for cross-support is an objective measure of demand for World Bank research. Each member of the research staff in DECRG is required to sell a minimum of 13 weeks of his or her time per year to other units, almost all of which are in operations. This arrangement exists in part to help ensure that researchers are exposed to the problems and challenges faced in the Bank’s operational settings, which in turn helps ensure the relevance of research. It also serves to make the operational staff aware of what their work gains from research. TTLs in operations typically come to the researchers with specific tasks, such as supervising new data collection or undertaking analytical work for an analytical product. 2.39 The survey analysis shows a positive relationship between the level of cross-support provided by DECRG to the VPUs and their familiarity with research; the level of cross-support provided to VPUs with a higher share of PhDs and of economists per capita is also higher (please see Table 2.10). Mean familiarity among operational staff is positively correlated with DECRG’s staffing, suggesting that the staffing of DECRG matters to familiarity with research among the 44 Bank’s operational staff. The sectors EP, EDU, HNP, POV, and SP account for 59 percent of staff in the research department, as compared to 15 percent for ARD, EM, TRN, and URB. Table 2.13: Summary Statistics on Perceptions of DEC Responsiveness across VPUs Responsiveness of DEC Availability of DEC Researchers Researchers for Cross-Support Vice Presidency (% rating 5+) (% rating 5+) Mean St. error Mean St. error Regional VPUs Sub-Saharan Africa 68.75 6.86 61.82 6.71 East Asia and Pacific 73.33 8.28 76.67 7.91 Eastern Europe and Central Asia 71.43 8.76 78.79 7.29 Latin America and Caribbean 66.67 9.31 73.08 8.91 Middle East and North Africa 71.43 12.39 76.47 10.54 South Asia 57.14 11.08 36.36 10.51 Network VPUs Finance and Private Sector 77.78 14.22 70.00 14.85 Human Development 78.57 11.25 66.67 13.94 Operational Policy 50.00 20.94 50.00 20.91 Poverty Reduction and Econ. Mgt. 86.36 7.51 76.47 10.54 Sustainable Development 61.54 13.84 50.00 12.81 World Bank Institute 100.00 0.00 75.00 22.18 Total: 70.95 2.93 66.93 2.96 Note: The totals include some other VPUs with insufficient sample sizes. Table 2.14: Summary Statistics on Perceptions of DEC Responsiveness across Sectors Responsiveness of DEC Availability of DEC Researchers Researchers for Cross-Support Sector (% rating 5+) (% rating 5+) Mean St. error Mean St. error Agriculture and Rural Development 70.59 11.36 75.00 9.94 Economic Policy 85.96 4.73 79.69 5.16 Education 60.00 11.26 52.63 11.76 Energy and Mining 37.50 17.60 50.00 16.23 Environment 75.00 15.74 75.00 15.72 Finance and Private Sector 63.64 10.55 58.33 10.33 Health, Nutrition, and Population 80.00 9.20 72.22 10.84 Poverty 90.48 6.59 85.71 7.84 Public Sector Governance 81.82 11.96 66.67 13.97 Social Development 20.00 18.39 28.57 17.53 Social Protection 66.67 16.16 66.67 16.13 Transport 37.50 17.60 50.00 18.15 Urban Development 55.56 17.03 50.00 16.23 Total 70.95 2.93 66.93 2.96 45 Conclusions 2.40 Although the Bank as a whole produces a large body of research, to date relatively little has been known about the demand for Bank research among operational staff. As background for this report, a survey of senior operational staff (grade GG and above) was conducted to measure their perceptions of Bank research. It found that most respondents view Bank research favorably, with over 65 percent in all VPUs attaching to it a value of 5 or higher on a scale of 1 to 10. While these results suggest room for improvement, especially for the remaining 35 percent, a closer look reveals significant heterogeneity in survey responses across regions, networks, and sectors. 2.41 Specifically, the survey found that staff working on poverty, human development, and economic policy tend to value and use research more than those in the more traditional sectors of Bank lending—agriculture and rural development, energy and mining, transport, and urban development. VPUs with higher shares of economists and PhDs (in any field) tend to value and use Bank research more, indicating an important role for mixture of operational skills in absorption of research. Also, sectors that use Bank research least tend to rely less on other (academic) research as well. Importantly, familiarity with research and its use improves with DECRG researchers’ involvement in operational work, and Bank staff expect their reliance on research to increase with time, perhaps because they are learning how useful it can be. Both VPUs and sectors that currently rely less on Bank research generally expect larger growth in usage. 2.42 The evidence suggests that stronger incentives for learning about World Bank research translate into higher investment in learning by staff, as reflected in their familiarity with research. Analysis indicates that about half of the gap between the mean familiarity scores for ARD, EM, TRN, and URB (on the one hand) and EP, EDU, HNP, POV, and SP (on the other) is accounted for by the difference in the value attached to research by staff working in these sectors. In this case, familiarity breeds respect: those staff who are more familiar with in-house research place a greater value on it and are more highly motivated to learn about it, which leads in turn to greater familiarity and more incentive to learn. 2.43 Most World Bank senior operational staff also believe the research department is successful in its role as the key in-house unit supplying relevant findings to Bank operations, and they support increasing funding to it rather than turning more to external sources. Most supportive of in-house research—and in favor of increased resources being devoted to it—are those who rely more on it and, especially, those who expect their reliance to increase with time. 2.44 Although the majority of respondents value research, the survey suggests scope for improvements. First, the Bank’s research program needs to continue to improve its alignment with operational needs and trends. Second, more work is needed to understand the factors, on both the supply and demand side, underlying the low levels of familiarity with research and the ineffectiveness of dissemination efforts in certain regions and networks. Third, interactions between researchers and operational staff could be strengthened, not only to increase the use of research in operations, but also to continue to improve the relevance of research for Bank operations. Future research reports plan to improve and repeat this survey to better understand these issues. 46 CHAPTER 3. RESEARCH-DRIVEN DATA AND SOFTWARE 3.1 The discipline of policy analysis is increasingly sophisticated and requires new theories, data, and tools to address emerging challenges. Over the years, the World Bank’s research department has been a major producer and disseminator of all three. It was the former Bank president Robert MacNamara who kick-started the process. Looking at the draft of the 1979 World Development Report (the first such report), MacNamara was shocked to see that only 17 developing countries had data on poverty and inequality, even though macroeconomic aggregates from national accounts were available for virtually all countries. McNamara told his research staff to go and collect the missing data. 3.2 Thirty years on, the resultant Living Standards Measurement Study (LSMS) has surveyed nearly 400,000 households in nearly 40 countries, with several countries being surveyed more than once. Multiple other data initiatives followed, many described in this chapter. Parallel to them was a major push by Bank researchers to produce methodological innovations on many fronts, as well. 3.3 Research and data work in the Bank have a long tradition of openness, with research findings debated publicly in a wide variety of external forums, and with Bank-generated datasets widely used and scrutinized in the broader development community. This openness is integral to ensuring quality and rigor in research and data outputs, and it underlies the Bank’s broader emphasis on an “Open Data and Open Knowledge� policy. Moreover, as analytical and computational tools used in development policy analysis have become more sophisticated, Bank researchers have been at the forefront of enabling and facilitating high-quality, policy-oriented analytical work by researchers in the developing world. 3.4 This chapter tells the story of how World Bank researchers ended up pioneering the compilation, construction, and collection of highly innovative datasets—datasets that have permitted new insights into key policy questions, often challenging the conventional wisdom on a topic. The chapter also tells a story about software development—how Bank researchers have developed software tools that have made accessible to analysts around the world complex methods and computational tools that otherwise might be too demanding in terms of technical training and computer software. DEC has not just been a pioneer in Open Data and Open Knowledge; it has been a pioneer of the idea of democratizing development economics—that is, empowering researchers in the developing world to do high-quality, policy-oriented analytical work themselves. The products it generates are a vital public good serving the development community. 3.5 This discussion focuses on the particular role of research in creating and improving the quality of data and analytical tools. Given this emphasis, it does not cover a number of highly visible institutional DEC data products, such as the World Development Indicators (WDI), which is the flagship Bank database for development produced by DECDG from established international sources. 47 Compilation, Collection, and Construction 3.6 Agreement is widespread that thoughtful data collection, guided by researchers who understand what types of information are most needed to address fundamental policy questions, is an important World Bank contribution to the field of development. This section discusses important research-based data products generated by the Bank. These products fall into three categories. Some are compilations of existing data series, while others are collections of primary data. The third category comprises datasets that are constructed by piecing together information from different sources and then performing some modeling. A vital characteristic of all three types of datasets is that they are valuable public goods: the same data can be used by many researchers and policy analysts to answer a broad range of development policy questions. 3.7 Below are described some of the key datasets produced by the Bank. A more comprehensive selection is shown in Table 3.1, which appears at the end of the section. Value of Compilations 3.8 The following are examples of the innovative and highly valued compilations of existing data series that have been produced by World Bank researchers. 3.9 Database of Political Institutions. The Database of Political Institutions (DPI) is the world’s most comprehensive compilation of data on political institutions, consisting of annual data on 100 variables for 100 countries over the period 1975–2010. The centrality of political incentives to development had long been accepted, but a conspicuous lack of data prevented researchers from quantifying their impact on development policies and outcomes. The DPI filled that gap in the political economist’s toolkit and made such work possible. 3.10 Scholars and practitioners have used the data in the DPI to examine how political institutions or ideology influence government decision making. For example, they have applied the data to analysis of competitive elections, the degree to which political parties are programmatic, the extent to which leaders control the ruling party, and the timing of elections to answer questions on subjects including the impact of elections on political budget cycles, the effects of political credibility on government incentives to provide public goods, and the role of political party institutionalization in attracting private investment. Google Scholar reports 1,596 citations to the article that introduced the DPI dataset as of January 2012. Its use is common across the social sciences, around the world, and in AAA and flagship reports across the Bank. 3.11 Worldwide Governance Indicators. During the 1990s it became clear that good governance was key to economic growth and development more broadly. What was missing was a way to measure it. The Worldwide Governance Indicators (WGI), compiled by the World Bank, have helped provide one. The dataset provides country-level measures of six dimensions of governance over the period 1996–2010: voice and accountability; political stability and absence of violence; government effectiveness; regulatory quality; rule of law; and control of corruption. The WGI are composite measures obtained by combining data from 30 existing 48 individual data sources that report the perceptions of and experience with governance of survey respondents and experts in the public and private sectors worldwide. 3.12 The WGI are used by policymakers and civil society groups as a tool for research, analysis, and advocacy regarding governance issues. They are also consulted by other aid donors and, most visibly, are used by the U.S. Millennium Challenge Corporation in determining country eligibility for assistance programs. The WGI are also frequently used by researchers, with several thousand citations in scholarly papers in the past five years. The WGI website has accumulated over 1 million visits since its launch in 2006, and the ten WGI-related Bank working papers have been downloaded over 100,000 times from the Social Science Research Network (SSRN) working paper archive. 3.13 Database of Financial Development and Structure. The Financial Development and Structure (FDS) database has sparked a large amount of cross-country analytical work, both inside and outside the Bank. The working paper that published this database is among the top 1 percent of papers listed in Research Papers in Economics (RePEc) in terms of the number of times cited elsewhere, and the version published in the World Bank Economic Review is among the top 10 articles of the journal in number of citations. The database provides statistics on the size, activity, efficiency, and stability of banks, equity markets, and bond markets across a broad spectrum of countries since 1960. As proper measurement is at the core of any rigorous analysis of drivers of financial development and their impact on growth and poverty alleviation, these indicators are important for analysts, researchers, and policymakers alike. The database also contains several indicators of financial globalization, including statistics on international bond issues, international loans, offshore deposits, and remittance flows, drawing on a wide array of primary sources to cover different dimensions of the financial system. 3.14 The FDS database has sparked further efforts within the Bank to collect financial sector indicators and benchmark countries, both on the global and regional levels. Through the years, many regional finance flagship reports in LAC, AFR, MENA, and ECA have used the indicators developed in this database in benchmarking their regions and countries. The Global Financial Development Report, along with a data analysis tool (Finstats), is expected to institutionalize and regularly update the FDS database. Collecting Primary Data 3.15 Data compilation relies on data being there to start with. The reason MacNamara saw so many gaps in the poverty and inequality data series in the 1979 WDR tables wasn’t that his staff hadn’t bothered to compile the data; the data simply did not exist. Most countries had never fielded a household survey, and those that had had not asked the necessary questions to estimate accurately a household’s consumption. Many asked about a household’s expenditure, but in developing countries in particular, where families grow their own food and engage in nonmarket transactions, market expenditures are a poor guide to consumption levels. To address this deficiency, the LSMS team set about designing a consumption questionnaire that could be used in household surveys in different settings, and they fielded it in multiple countries. Other 49 organizations subsequently embraced the philosophy of the consumption module, thanks in part to a three-part volume the World Bank produced explaining the idea and how to implement it. Another key motivation for the development of LSMS was the inadequacy of the pre-existing surveys for most policy applications, such as studying the incidence and impacts of taxes and social and human development programs. In response to all these needs, a new generation of household surveys emerged, and, 30 years later, nearly 400,000 households in nearly 40 countries have been surveyed using a uniform methodology. 3.16 Living Standards Measurement Survey. Once the initial goal of LSMS of getting a good fix on household consumption was fulfilled, modules began to be added on anthropometrics, education, employment, health, transfers and other non-labor income, and housing. The LSMS was soon being used by researchers to explore a whole host of questions in human development and labor markets. The LSMS also started collecting community data—not just on local prices, which are key to adjusting and understanding consumption patterns, but also on local infrastructure, travel time to the closest market, and the availability of local services, such as education and health care. Sometimes the module was expanded to include visits to schools and clinics, as well as to key informants. Optional modules added in some countries covered such topics as agriculture, the environment, fertility, and credit and savings. 3.17 The LSMS team is currently developing new survey methods and tools to improve the quality and usefulness of data and increase the efficiency of their collection. It is also updating existing tools and creating new ones to expand the dissemination of literature, knowledge, and results in the area of survey methodology in developing countries. A variety of experiments is being undertaken in the areas of consumption, income, labor, subjective welfare, inequalities of opportunity, disability, financial services, risk and vulnerability, migration, infrastructure, and gender to try out different ways of obtaining information on such topics. 3.18 Technological innovations are also in progress. The research group is developing new Computer-Assisted Personal Interview System (CAPI) software for tablet computers like the iPad. Access to CAPI on tablet computers will enable data entry in the field that allows errors and inadmissible values to be caught, as well the addition of time stamps, GPS coordinates, and audio and video recordings of interviews. 3.19 Parallel to these developments is a $19 million project funded by the Bill and Melinda Gates Foundation to improve household data on agriculture in Africa. In recognition that existing agricultural data in the region suffers from inconsistent investment, institutional and sectoral isolation, and methodological weakness, the LSMS Integrated Surveys on Agriculture (LSMS- ISA) project collaborates with the national statistics offices of its seven partner countries in Sub- Saharan Africa to design and implement systems of multi-topic, nationally representative panel household surveys with a strong focus on agriculture. Its primary objective is to foster innovation and efficiency in statistical research on the links between agriculture and poverty reduction in the region. 3.20 The LSMS-ISA project is committed to providing timely and open access to all microdata it produces. The data collected in each project country are fully cleaned, documented, 50 and made publicly available within 12 months of the completion of each survey round. Data for Nigeria and Tanzania are already available. 3.21 In addition to conducting surveys of households in developing countries, World Bank researchers, in an ambitious exercise, set out to measure the quality of medical care in public and private ambulatory care clinics. This data initiative—which has generated freely available survey instruments as well as datasets that will ultimately cover several countries—has collected vital and sometimes shocking information. The tools are increasingly being used in operations in economic and sector work (ESW). 3.22 Investment Climate and Firm Surveys. Frustrated by the limits of available data, Bank researchers began in 2001 to survey firms in multiple countries systematically using a common questionnaire. Until then, only a few countries had publicly available firm-level survey data. Now, 11 years later, mainstreamed and housed in FPD, the initiative has built up comparable data on over 130,000 establishments across 125 countries. The surveys include data on firm characteristics (such as age, sector, ownership, and location), firm performance (such as profitability, productivity, employment growth, investment, and innovation), and measures of the investment climate in which the establishments operate. This was one of the first systematic attempts to augment subjective indicators of the investment climate with more quantitative and comparable measures across a range of indicators of the business environment. Thus, for example, in addition to asking whether electricity is a constraint on the operation and growth of an establishment, surveys also collect information on the frequency and length of outages and the production lost due to them. The instruments have also been tailored to respond to changing conditions and to monitor the impact of the current financial crisis. For example, six countries in ECA have run modules of the surveys every six months to follow the differential impact of the financial crisis across firms. 3.23 While researchers developed the questionnaire and the sample design and analyzed the effects of investment climate on firm performance, the Bank’s operational units popularized the use of the data and its Investment Climate Assessments and the monitoring of investment climate indicators. The focus now is on building up panel datasets over time. Today many countries have had second- and third-round surveys completed, allowing the analysis of variations in investment climate conditions. In addition to being used in a long list of research papers, the data are widely cited in FPD and PREM regional flagships addressing issues of competitiveness, productivity, and employment, as well as in World Development Reports. 3.24 Purchasing Power Parities. Market exchange rates, which are often used to compare prices across countries, give misleading comparisons because they do not reflect international differences in purchasing power. Purchasing Power Parity (PPP) is a form of exchange rate that takes into account the cost of a common basket of goods in two countries being compared. For example, the PPP for the Indian rupee against the U.S. dollar is defined as the number of rupees needed to buy in India the same amount of goods and services as one dollar would buy in the United States. National incomes also have to be converted into a common currency before comparisons can be made. Recognition of the deficiencies associated with market exchange rates thus led to use of PPP as a currency converter to compare incomes across countries. But 51 calculating PPP requires enormous amounts of detailed consumption and production data at the country level. 3.25 To meet this challenge, the International Comparison Program (ICP) was established in 1968 as a joint venture of the United Nations Statistical Division, the International Comparisons Unit of the University of Pennsylvania, and the World Bank, with financial contributions from the Ford Foundation. Started as a modest research project, its ultimate goal was to set up a regular program of worldwide PPP-based comparisons of GDP. 3.26 Comparisons of final expenditure on GDP have been carried out seven times since 1970. The number of countries originally included was 10; the 2011 round of the ICP, currently underway, includes 197 economies. The World Bank hosts the global office of the ICP and continues to draw upon the knowledge of researchers inside and outside the Bank to guide its development. 3.27 In the 2005 round, the World Bank, together with members of the technical advisory group of senior practitioners and theorists that oversees this work, produced important new research on poverty PPPs and introduced a new approach for linking regional into global PPPs. The 2011 round has completed research in additional areas, including the introduction of a new approach for computing construction sector PPPs and research on productivity adjustment for the ICP. Research is still in progress on applications of PPPs that include identifying a core set of data required for poverty research purposes, extrapolating PPPs between benchmark years, and developing approaches for computing sub-national PPPs. 3.28 Bank Regulation and Supervision. Recent financial crises have pushed bank regulation and supervision and too-big-to-fail and deposit insurance issues to the center of debates on regulatory policy in both developing and industrialized countries. In both areas, World Bank researchers have been doing significant work, often questioning conventional wisdom and prompting the development of new datasets. The Bank Regulation and Supervision Database is a unique cross-country database on bank regulation and supervision in over 150 countries. Right incentives require ensuring the availability of accurate and timely information, and analysis of the Bank Regulation and Supervision data has shown the importance of preserving and providing incentives to harness market forces to complement the limited capacity of official supervision. These findings have influenced the Bank’s policy advice on the Basel II Capital Accord. 3.29 Now the Dataset on Deposit Insurance, a subset of the Bank Regulation and Supervision Database, has led to research that has questioned the adoption of deposit insurance by highlighting the potential costs of explicit schemes—lower market discipline, higher financial fragility, and lower financial development—in countries where complementary institutions are not strong enough to keep their costs under control. Guided by this research, the Bank recommended against an early adoption of deposit insurance in Russia and China, among many others, and encouraged the policymakers to focus on more pressing financial sector reforms instead. 3.30 Global Financial Inclusion Indicators (Global Findex). Research often prompts data generation by exposing the limitations and gaps in existing datasets. Financial inclusion is one 52 area where such gaps have been found. The Policy Research Report, “Finance for All? Policies and Pitfalls in Expanding Access,� synthesized a large body of evidence and documented the extent of financial exclusion around the world. It showed the importance of access to financial services for growth, equity, and poverty reduction and discussed policy interventions and institutional reforms that can improve access for underserved groups. Lack of access to finance is often the critical driver of persistent income inequality as well as slower growth, and the report illustrated that less than half the population across the developing world has an account with a financial institution; in Sub-Saharan Africa, the figure is less than 20 percent. 3.31 One important message of the report was how limited data on financial inclusion are. The report received significant policy interest from and coverage by The Economist, was featured in the U.N. Year of the Microcredit, and fed into the G-20 discussions in Korea. It also facilitated greater data collection efforts on access and inclusion by CGAP and by IMF, which has now institutionalized the supply-side inclusion indicators. More recently, the research department—in partnership with the Gallup Organization and financed by a large ten-year grant from the Gates Foundation—has started collecting demand-side data. The Global Financial Inclusion Indicators (Global Findex) will improve the understanding of how individuals around the world save, borrow, and make payments. These new indicators, expected to be released at annual meetings in spring 2012, will also be central to the financial inclusion agenda of the Mexican G-20 discussions. Overall, the World Bank’s research and data work in this area have catalyzed much greater policy interest, also contributing to the formation of a new global practice for financial inclusion in the newly restructured FPD Network. Constructing Datasets 3.32 The Database on Political Institutions and the World Governance Indicators lie at one extreme of database creation, and the LSMS and investment climate surveys lie at the other. The former are highly valued compilations of existing data series, while the latter are collected through questionnaires and observation in thousands of households and firms all over the developing world. Between these two extremes lies another type of dataset: one that is constructed by piecing together information from different sources and then performing some modeling. The following are examples of datasets constructed by World Bank researchers. 3.33 Dollar-a-Day. The World Bank’s highly visible global poverty monitoring efforts— including the famous “dollar-a-day� measures—have been based in the research department and currently draw on 850 household surveys for 130 developing countries. The Bank’s leadership in this area is widely recognized. For example, The Economist magazine stated in its issue of March 3–9, 2012, that “the best estimates for global poverty come from the World Bank’s Development Research Group,� and proceeded to describe in detail the latest estimates. Though it draws on data sources and expertise throughout the World Bank, from other regional banks (through the International Comparison Program), and from developing countries directly, it is no accident that this work is housed in the research department. Researchers have the analytical skills and independence crucial to establishing the credibility of these numbers, which is also supported by scholarly peer review. The papers documenting the work are extensively reviewed within and outside the Bank before being submitted to leading scholarly journals, where they are subjected 53 to further anonymous peer review by academic experts. For example, the methodology used for the latest estimates was published in one of the leading economics journals, the Quarterly Journal of Economics, after thorough anonymous peer review that improved the paper. The research department also makes the core data openly available at zero cost through its web-based data tool, PovcalNet (discussed further in the next section). 3.34 The Overall Trade Restrictiveness Indices. The Overall Trade Restrictiveness Indices (OTRI) have been widely used by researchers and trade negotiators and to inform World Bank publications, such as the Global Monitoring Reports. The data, which are constructed around the relevant economic theory, take into account the structure of trade policy (both tariff and nontariff barriers), trade flows, and trade elasticities of each country. This database is used to estimate trade protection by developing and developed countries. The results highlight the differences in trade protection among low-income, middle-income, and least developed countries and reveal that while poor countries have more restrictive policies, they face higher trade barriers on their exports. 3.35 Temporary Trade Barriers Database. The Temporary Trade Barriers Database (TTBD) is an outgrowth of the Global Antidumping Database (GAD). Detailed data are collected and freely disseminated on the use by more than 30 national governments of the most common temporary trade barriers permissible in the World Trade Organization (WTO) system: antidumping duties, countervailing duties, and safeguard actions. This data collection effort, which began with the free internet publication of the GAD in 2005, is the first attempt to use original national government documentation to organize information on products, firms, the investigative procedures, and the outcomes of the historical use (since the 1980s) of these policy instruments across most of the WTO system’s users. It is still the only electronically available database of the use of trade remedies by WTO member economies. Not only have the GAD data been used extensively in academic and policy research; their time series nature was frequently relied upon to monitor the spread of new trade barriers implemented during the global economic crisis of 2008–2009. Since the beginning of 2009, detailed updates to the GAD have been freely provided on the internet on a quarterly basis. 3.36 Global Bilateral Migration Database. The Global Bilateral Migration Database consists of five 226-by-226 bilateral matrices of migration stocks by gender for each decade over the period 1960–2000. Raw data from over 1,000 national censuses and population registers are used and various numerical exercises performed to fill any gaps. Because the data covering migrant destinations in these thousands of separate national sources are not directly comparable—and therefore not suitable for analysis—a major effort has been made to harmonize the matrices and verify the reliability of the resulting estimates. 3.37 Legal Data. A key question about judicial involvement in policymaking is the extent to which the poor are able to benefit from rights-based litigation. A number of scholars have suggested that social and procedural barriers give the rich and powerful advantages that likely prevent the great majority of poor people from making claims in courts. With empirical work regarding the legal system still in its infancy, testing whether the legal system displays a pro-rich bias is a challenge. To address it, World Bank researchers undertook the considerable task of coding the contents of court decisions relating to public interest litigation by type of claim and 54 social class of the claimant, beginning with cases that reached the Indian Supreme Court during the past forty years. These data have been used for research and have been cited in a number of media outlets and blogs. Table 3.1: Selected Examples of Research-Driven Databases A. Compiled Databases Database Description Database of Political The Database of Political Institutions is the world’s most Institutions comprehensive compilation of data on political institutions, consisting of annual data on 100 variables for 100 countries over the period 1975–2010. Worldwide Governance The Worldwide Governance Indicators Database reports the Indicators perceptions of and experience with governance of survey respondents and experts in the public and private sectors worldwide. Financial Development and The Financial Development and Structure Database provides statistics Structure Database on the size, activity, efficiency, and stability of banks, equity markets, and bond markets across a broad spectrum of countries since 1960. Agricultural Trade Distortions The Agricultural Trade Distortions Database is a compilation of Database detailed country-level information on taxes, tariffs, and other trade distortions in agriculture for the period 1955–2007. Services Policy Restrictiveness The Services Policy Restrictiveness Database contains information on Database policy and regulatory measures affecting international trade in services for a total of 102 countries, of which 78 are developing countries, and 24 OECD countries. B. Collected Databases Database Description Living Standards The Living Standards Measurement Survey is a multi-dimensional Measurement Survey survey of households to collect data on consumption, income, and various other indicators that help in assessing household welfare and living conditions. Investment Climate and Firm Investment Climate and Firm Surveys systematically survey firms in Surveys multiple countries using a common questionnaire to augment subjective indicators of the investment climate with more quantitative and comparable measures across a range of indicators of the business environment. Purchasing Power Parities Purchasing Power Parity is a form of exchange rate that takes into account the cost of a common basket of goods and services. Bank Regulation and The Bank Regulation and Supervision Database is a unique cross- Supervision Database country database on bank regulation and supervision in over 150 countries. Global Financial Inclusion Global Financial Inclusion Indicators are new indicators intended to Indicators (Global Findex) gather data on the access to financial services and the financial inclusion of individuals around the world. 55 Quantitative Service Delivery A quantitative service delivery survey assesses health and education Survey services in country-specific contexts in a way that facilitates cross- country comparisons. Public Expenditure Tracking A public expenditure tracking survey tracks government spending as it Survey flows from the finance ministry down to grass-roots schools and health facilities. Absenteeism Surveys Random unannounced surveys of schools and health facilities determine rates of staff absenteeism in an effort to discover why these institutions often fail to deliver results. Maximum Residue Level The Maximum Residue Level Database collects maximum pesticide Database residue levels for agricultural products in a number of developed and developing countries. C. Constructed Databases Database Description Overall Trade Restrictiveness The Overall Trade Restrictiveness Indices provide trade protection by Indices developing and developed countries by taking into account the structure of trade policy (both tariff and nontariff barriers), trade flows, and trade elasticities. Temporary Trade Barriers The Temporary Trade Barriers Database, an outgrowth of the Global Database Antidumping Database, collects and freely disseminates detailed data on temporary trade barriers permissible in the World Trade Organization system. Global Bilateral Migration The Global Bilateral Migration Database consists of five 226-by-226 Database bilateral matrices of migration stocks by gender for each decade over the period 1960–2000. Legal Data Legal Data provide information on court decisions relating to public interest litigation by type of claim and social class of the claimant. Environmental Data The World Bank’s environmental datasets are constructed by piecing together information from different sources, including information gleaned from maps and satellite images, and then performing some modeling to address various questions about the environment. Exporter Dynamics Database The Exporter Dynamics Database, which includes statistics on export growth and dynamics in the agricultural, mining, and manufacturing sectors, is constructed using exporter-level customs transaction data as an input for 30 or more developing countries and a handful of developed countries. 56 Democratizing Development Economics through Software 3.38 Research plays a valuable role in expanding and improving the methodologies and tools routinely employed by policymakers and analysts engaged in the World Bank’s country analytical work and by those in client countries. As the field of development policy analysis has progressed, these analytical tools have become increasingly sophisticated and complicated. Often, lack of necessary technical and programming skills can be a barrier to entry for analysts who want to apply “best practice� tools to their work. The World Bank has devoted substantial resources—especially in the past five years—to developing software geared to helping researchers and analysts around the world take advantage of the data and methods available to them. The focus here is on computational tools designed to implement complex analysis, as opposed to software designed to facilitate data retrieval and display. ADePT Software Platform 3.39 ADePT, which is a multi-module software platform for automated economic analysis, provides easy access to highly sophisticated quantitative methods, many developed by Bank researchers who are global leaders in their fields. An attractive, easy-to-use, menu-driven interface allows ADePT users to specify the tables and charts they would like to produce from the household survey data they have loaded. The user might want to see, for instance, who benefits most from government health spending in India, or how much higher poverty is in rural China than in urban China, or how well Mexico’s conditional cash transfer program targets the poor. One might also unravel the causes of inequalities in child malnutrition in South Africa—to what extent they are due to inequalities in access to water and sanitation, rather than, say, inequalities in parents’ education. 3.40 In contrast to standard statistical and econometric software, ADePT tells the user what tables and charts can be produced with the dataset and the variables in it. The user simply has to choose from the available tables and charts and indicate which variables to use in the analysis. ADePT does the rest, producing print-ready tables and charts that require no further formatting. 3.41 ADePT is, in effect, an easy-to-use multilingual interface (the user can choose one of 15 languages) that “talks to� Stata, the statistical and econometric software that sits under it. Through a generous agreement with the Stata Corporation, the Stata software is bundled with ADePT at zero cost to the user. ADePT thus enables users to use complex quantitative methods without the need to do the necessary programming in Stata, and gives them the power of Stata at no cost. For analysts and researchers in the developing world, both are big pluses. Many have not had the opportunity to undertake advanced studies in applied quantitative methods, and software like Stata is often unaffordable. But ADePT also makes accessible to non-economists quantitative methods developed by economists. Many who attend the training events on the ADePT health module, for instance, have backgrounds in medicine, public health, and epidemiology, not economics. In fact, many have been civil servants and program managers from other international agencies, keen to take advantage of the easy-to-use interface to help them make informed decisions about their organizations’ work programs in the field. 57 3.42 ADePT offers other benefits, too. Even seasoned Stata users can make programming mistakes, and different analysts may go about computing the same statistical results in different ways. ADePT eliminates the risk of mistakes—the software is tested extensively before release—and avoids the possibility of differences in calculations resulting from differences in the programming approach used. Perhaps the biggest attraction of all, however, is the time saved by researchers and analysts in producing the tables and charts they need. Instead of devoting hours to learning how to implement a method written up in a journal article, the analyst can focus on careful data preparation and on the interpretation of the results and their implications for policy. 3.43 ADePT—an acronym for Automated DEC Poverty Tables—began with a focus on poverty analysis. The original poverty module includes powerful new techniques, such as the growth-incidence curve, growth-inequality, regional decompositions, and stochastic dominance analysis, all developed by Bank researchers. ADePT now has modules for education, health, gender, inequality, labor, and social protection, as well. Some of these—notably the health and inequality modules—reflect years of research by Bank staff. Others have been developed in partnership with Bank experts in the relevant topic from the relevant networks. Recent developments include the construction of a module that allows the user to simulate the impact of economic shocks, farm subsidies, cash transfers, and other policy instruments on poverty, inequality, and labor market outcomes. 3.44 A recognition of ADePT as a valuable tool for economic analysis drew attention to the product from other international development agencies. ILO, FAO, UNISEF, and EU expressed interest in developing specialized modules for their clients. Among the recent examples of cooperation between the World Bank and these agencies are the ADePT Food Security Module (FSSM), jointly developed with UN FAO, and the Labor Market Indicator (LMI) module, developed with ILO. The ADePT FSSM implements standard statistical methodology and procedures for the assessment of hunger reduction at the global, regional, and country levels. The module is designed to analyze the underlying causes of hunger and malnutrition, to monitor progress toward hunger reduction, and to raise awareness about global hunger issues. It will allow the governments of more than 100 member-countries of FAO to analyze their own data to assess the state of food security in their countries. 3.45 The ADePT LMI was developed jointly with the ILO Employment Trends Unit. The module, which uses data from labor force surveys and integrated household surveys, allows users to analyze issues of employment and social protection and to assess the effectiveness and impact of individual measures to protect the most vulnerable population groups. ILO plans to disseminate this module extensively among the member-countries and has designed several training courses to promote its use for labor market analysis in developing countries. The development of the ADePT FSSM and LMI modules extends and promotes to other international agencies the World Bank initiative for openness. 58 PovcalNet 3.46 While the World Bank’s dollar-a-day poverty line has been widely used in the U.N. system and beyond, and the Bank reports poverty estimates for other lines as well, researchers and analysts need to be able to use their own poverty lines. PovcalNet allows users to conduct independent poverty analysis by applying their own poverty lines to the same household surveys used in the World Bank’s dollar-a-day poverty calculations and using the same computational methods. Users can undertake calculations on the household survey datasets—or parametric representations of them—that the Bank uses, but cannot retrieve or otherwise analyze the data; the software thus enables researchers and analysts to use the survey results while ensuring the Bank honors its agreements with national governments not to distribute the raw data. 3.47 PovcalNet currently has income and/or consumption distributional data from more than 800 household surveys, spanning 1979–2010 and 126 developing countries. More than 1.2 million randomly sampled households were interviewed in these surveys, representing 96 percent of the population of developing countries. The database is updated several times a year as new survey data become available, and a major reassessment of progress against poverty is made about once every three years. PovMap 3.48 Poverty maps provide spatial and visual representations of the distribution of poverty in a country. They are of greatest use to policymakers and researchers when they are finely disaggregated—that is, when they represent small geographic units, such as cities, towns, or villages. Many policymakers in developing countries use poverty maps when planning public investments in education, health, sanitation, water, transport, and other sectors. Social funds often use them because geographically targeted investments are thought to reach many poor citizens and have spillover effects that enhance productivity in depressed areas. 3.49 Unfortunately, almost all household surveys are too small to be representative at such fine levels of disaggregation, and few census (or other large) datasets contain the detailed information needed to calculate monetary indicators of poverty or inequality. In general, the prohibitive cost of collecting a lot of information for a very large sample produces a tradeoff between the number of observations in a dataset and the information content. One way to overcome this obstacle is to apply econometric techniques to combine sample survey data with census data to predict consumption-based poverty indicators using all households covered by the census. This method produces estimates of such familiar indicators as the poverty headcount or the Gini coefficient, as well as an indication of the degree of statistical precision of these estimates, for small administrative units. 3.50 A team of researchers in the Bank has developed a methodology along these lines and implemented it in more than 70 countries. The team provides technical assistance and capacity building, as well as a powerful, platform-independent and freely available software program 59 called PovMap, to facilitate the production of such poverty maps. PovMap, which provides computational solutions at all stages of poverty mapping activities, uses a proprietary data engine to ensure the speedy processing of census data. Its functions include data processing, distributional comparison, regression analysis, univariate analysis, correlation analysis, cross- tabulation, and simulation. iSimulate and the GEP Forecasting Model 3.51 The research and development of iSimulate, an open web platform for performing collaborative economic simulations, has allowed researchers to strengthen their forecasting models and their ability to collaborate with regions, easily tapping into the expertise, monitoring, and econometric work of countries around the world. The forecasting model uses a global macro model covering more than 150 countries to produce its forecasts for the Global Economics Prospects (GEP) report and includes both national income accounts and balance of payments components. iSimulate connects the World Bank’s three pillars of openness—open tools, open data, and open knowledge. Its expansion would cover other economic models and tools— MAMS, GIDD, the 1-2-3 model, and ENVISAGE—that are widely used by economists and policymakers. World Integrated Trade Solution 3.52 The World Integrated Trade Solution (WITS) is both a data initiative and a software initiative. A collaborative venture with UNCTAD, WITS allows users to retrieve and analyze information from four different databases on trade and trade barriers: the U.N. Statistical Division’s COMTRADE database of trade flows; UNCTAD’s database of tariffs and nontariff barriers; WTO’s applied tariff rates and tariff bindings; and the Agricultural Market Access Database Consortium’s database of agricultural trade barriers. The Bank’s contribution has been to develop the software, while UNCTAD’s role has been to collect the data. The objective of the WITS project is to provide the best possible information on trade and trade policies to its users and the ability to transform these data in ways that are valuable to them. The two major groups of users are trade policymakers and their advisers, and policy and research/country economists in the World Bank and other institutions. 60 Table 3.2: Select Examples of Research-Produced Software and Analytical Tools Software Description ADePT Software Platform ADePT is a multi-module program that provides easy access to highly sophisticated quantitative methods, many developed by Bank researchers who are global leaders in their fields. PovcalNet PovcalNet allows users to conduct independent poverty analysis by applying their own poverty lines to the same household surveys used in the World Bank’s dollar-a-day poverty calculations and using the same computational methods. PovMap PovMap applies econometric techniques to combine sample survey data with census data to predict consumption-based poverty indicators using all households covered by the census. iSimulate iSimulate is an open web platform for performing collaborative economic simulations, allowing researchers to strengthen their forecasting models and their ability to collaborate with regions. 1-2-3 The 1-2-3 model is a simple, computable general equilibrium model that uses minimal national accounts data to analyze a wide range of country economic policy issues. Environmental Impact and The Environmental Impact and Sustainability Applied General Sustainability Applied General Equilibrium model (ENVISAGE) is the World Bank’s global Equilibrium economic model for constructing long-term scenarios and analyzing structural policies. Global Income Distribution Dynamics Global Income Distribution Dynamics is the first global CGE- microsimulation model that measures the effects of economic policies on poverty and on the distribution of welfare among individuals and households. Maquette for MDG Simulations Maquette for MDG Simulations is an extended, dynamic, country-level applied general equilibrium model that incorporates development outcomes in education, health care, and other areas. World Integrated Trade Solution The World Integrated Trade Solution allows users to retrieve and analyze information from five databases on trade and trade barriers. Conclusions 3.53 In recent years, the World Bank has been producing ever more data and tools to facilitate development analysis and policymaking. In pioneering a strategy of collecting data from scratch where information was lacking, Bank researchers have enabled the exploration of key policy questions that could not be answered using existing datasets, making it possible to compare countries and track progress over time and to increase monitoring activities substantially. LSMS, for example, provides a multi-dimensional assessment of household welfare, as well as allowing evaluation of the effect of various government policies on the living conditions of the population. Similarly, ICP data allow the setting of international poverty lines with common purchasing power, comparisons of relative sizes of economies, and the allocation of quotas and funds across countries and regions. 61 3.54 Research has also played a valuable role in expanding and improving the toolkit routinely employed by policymakers and analysts engaged in the World Bank’s country analytical work and by those in client countries. In response to the Open Bank Initiative, the Bank has increased its involvement in the facilitation of applied economic research, devoting substantial resources to developing software products geared to helping researchers and analysts around the world take advantage of the data and methods available to them. ADePT is an excellent example of a computational tool invented by the Bank to make existing analytical tools of inequality and poverty measurement easy to implement. PovcalNet allows users to conduct independent poverty analysis by applying their own poverty lines to the same household surveys used in the World Bank’s poverty calculations and using the same computational methods, while also ensuring the Bank honors its agreements with national governments not to distribute the raw data. The research and development of iSimulate, an open web platform for performing collaborative economic simulations, has allowed researchers to strengthen their forecasting models and their ability to collaborate with regions, easily tapping into the expertise, monitoring, and econometric work of countries around the world. 3.55 By leading an expansion of the volume, quality, and accessibility of development data and providing tools, World Bank research is providing a platform on which developing country researchers and policymakers can do high-quality, policy-oriented analytical work themselves. This has strengthened the foundation for development knowledge and well-informed policymaking. 62 ANNEX A. RESEARCH PROGRAMS 1. The World Bank’s research program is guided by themes and topics that are crucial for economic development. Large multiyear projects analyze issues to provide strategic direction on critical problems facing Bank operations as well as its clients. Each centers on a theme, giving coherence to the research and its outputs. During FY 2009–2011, over 360 research projects covering a wide array of topics were being implemented each year. Economic management, environment, and finance and private sector themes accounted for nearly half of the projects. Other important themes include human development, trade and international integration, and social protection. 2. This section presents a synopsis of Bank research by thematic groups and highlights some significant contributions under each theme. A.1 Agriculture and Rural Development Research Program 3. Research on agriculture and rural development is critical to the Bank’s mission because poverty remains largely a rural phenomenon, with 75 percent of the world’s poor living in rural areas and the majority depending on agriculture for their livelihoods. The performance of the agricultural sector is also critically important because of the impact of food prices on the cost of living and food security of the urban poor. Themes 4. The research program on agriculture and rural development covers five broad themes: agricultural productivity and factor markets; rural infrastructure, governance, and public goods; agriculture and the environment; rural-urban migration; and price incentives, trade and vulnerability. Highlights Agricultural productivity and factor markets 5. The Asian green revolution transformed Asian agriculture, and contributed to the dramatic progress of many Asian economies in economic growth and poverty reduction. A key question is whether, and if so how, a green revolution can be achieved in Africa. A recent, large- scale study of African agriculture concludes that African agriculture has considerable potential, but that these benefits may be more difficult to achieve because of the much greater diversity of African agriculture. 9 An important element of an African green revolution—like its Asian predecessor—is likely to be improvement in productivity and yields in rice, since many of the 9 Larson, Donald F., and Keijiro Otsuka. Forthcoming. An African Green Revolution: Finding Ways to Increase Productivity. New York: Springer. 63 practices that proved successful in Asia can be applied where yields are currently low in Africa. 10 6. Uncertainty about land use rights may greatly inhibit agricultural investment and productivity improvement. But improving land governance is challenging. To address these challenges, the Land Governance Assessment Framework (LGAF), developed by the World Bank in close partnership with other multilateral and bilateral institutions, is a framework to assess land governance at the country level, in a participatory process that can draw together existing evidence from a wide range of sources. 11 This builds on earlier studies that show the potential for substantial gains from low-cost approaches to securing land rights. 12 7. Rising agricultural prices and concerns about food security have prompted a wave of interest in international investment in agricultural land. The amount of uncultivated, unforested, unprotected land suitable for rainfed agriculture is estimated at around 445 million hectares, and almost 60 million hectares worth of deals—more than 70 percent of them in Africa—were announced in 2009. With gaps in property rights and institutions, and “investors� often targeting countries with weak land governance, locals and vulnerable groups often lost out. 13 8. Improving credit to the poor is a major challenge and the Program Initiatives for Monga Eradication (PRIME) was introduced in Bangladesh to address concerns that regular microfinance does not adequately address hard-core poverty. Besides providing loans, it offers extension and training services. A quasi-experimental study found that PRIME is more effective than regular microfinance in reaching the very poor and reducing seasonal deprivation and extreme poverty. 14 A critical problem for strengthening provision of rural finance is the problem of failure to repay loans. A study of group loans in 300 Indian villages found that regular 10 Larson, Donald F., Keijiro Otsuka, Kei Kajisa, Jonna Estudillo, and Aliou Diagne. 2010. “Can Africa Replicate Asia’s Green Revolution in Rice?� Policy Research Working Paper 5478, World Bank, Washington, DC. Nakano, Yuko, Ibrahim Bamba, Aliou Diagne, Keijiro Otsuka, and Kei Kajisa. 2011. “The Possibility of a Rice Green Revolution in Large-Scale Irrigation Schemes in Sub-Saharan Africa.� Policy Research Working Paper 5560, World Bank, Washington, DC. Melinda Smale, Derek Byerlee, and Thom Jayne. 2011. “Maize Revolutions in Sub-Saharan Africa.� Policy Research Working Paper 5659, World Bank, Washington, DC. 11 Deininger, Klaus, Harris Selod, and Anthony Burns. 2012. The Land Governance Assessment Framework: Identifying and Monitoring Good Practice in the Land Sector. Agriculture and Rural Development Series. Washington, DC: World Bank. 12 Deininger, Klaus, Daniel Ayalew Ali, Stein Holden, and Jaap Zevenbergen. Forthcoming. “Rural Land Certification in Ethiopia: Process, Initial Impact, and Implications for the Other African Countries.� World Development. (Based on Policy Research Working Paper 4218, World Bank, Washington, DC.) 13 Deininger, Klaus, and Derek Byerlee, with Jonathan Lindsay, Andrew Norton, Harris Selod, and Mercedes Stickler. 2011. Rising Global Interest in Farmland: Can It Yield Sustainable and Equitable Benefits? Washington, DC: World Bank. 14 Khandker, Shahidur R., M. A. Baqui Khalily, and Hussain A. Samad. 2010. “Seasonal and Extreme Poverty in Bangladesh: Evaluating an Ultra-Poor Microfinance Project.� Policy Research Working Paper 5331,World Bank, Washington, DC. 64 monitoring and audits, high repayment frequency, and having group savings deposited with the lender all significantly increase repayment rates. 15 Rural infrastructure, governance, and public goods 9. While it is clear that more engaged communities can achieve higher levels of cooperation, little is known about how best to secure this engagement. A comprehensive study of the evidence finds that greater community involvement tends to improve resource sustainability and infrastructure quality. However, this frequently requires creation of local management capacity, with clear mechanisms for downward accountability, and communities need to benefit from the resources they manage. The report has potentially major implications for the way the Bank and other donors manage their portfolios of Community-Driven Development projects. 16 10. Where governance is weak and contract enforcement difficult, communities tend to develop informal contractual solutions. This is exemplified by the pervasive marriage custom in rural Pakistan of watta satta, a bride exchange between families. The likelihood of estrangement, domestic abuse, and other problems, is found to be lower in watta satta arrangements than in conventional marriages. 17 11. The World Bank is a major provider of rural roads but detailed knowledge about the payoffs from improved roads is unavailable. A detailed study of a proposed investment in rural roads in Madagascar found that it would raise the incomes of the most isolated by around 50 percent, largely due to increased opportunities for nonfarm earnings near towns. 18 Another study found that a road project in Bangladesh reduced poverty by around 5 percentage points in the first five years. 19 These studies and others in this vein show that the costs of isolation extend far beyond the transport costs of sending exports out and bringing imports in. 20 12. Rural electrification can bring potentially large benefits to households and the communities they live in. A study of World Bank–sponsored rural electrification projects in 15 Deininger, Klaus, and Yanyan Liu. 2009. “Determinants of Repayment Performance in India Micro-credit Groups.� Policy Research Working Paper 4885, World Bank, Washington, DC. 16 Rao, Vijayendra, and Ghazala Mansuri. Forthcoming. Localizing Development: Does Participation Work? Policy Research Report. Washington, DC: World Bank. 17 Jacoby, Hanan G. , and Ghazala Mansuri. 2010. “Watta Satta: Bride Exchange and Women’s Welfare in Rural Pakistan.� American Economic Review 100(4): 1804–25. 18 Jacoby, Hanan, and Bart Minten. 2009. “On Measuring the Benefits of Lower Transport Costs.� Journal of Development Economics 89: 28–38. 19 Khandker, Shahidur R., Zaid Bakht, and Gayatri B. Koolwal. 2009. “The Poverty Impact of Rural Roads: Evidence from Bangladesh.� Economic Development and Cultural Change 57(4):685–722. 20 Fafchamps, Marcel, and Forhad Shilpi. 2009. “Isolation and Subjective Welfare: Evidence from South Asia.� Economic Development and Cultural Change 57(4): 641–83. 65 Vietnam concludes that the benefit-cost ratio of electrification was over four. 21 In Bangladesh, a study using impact evaluation techniques finds that the benefit from rural electrification may be as high as 30 percent of households’ initial incomes, and that the benefits substantially exceed the costs. 22 Agriculture and the environment 13. Research on the impact of climate change in Africa suggests that the impacts may be quite diverse, with different impacts between agro-ecological zones, with currently productive areas such as dry/moist savannah being more vulnerable to climate change while humid forest or sub-humid zones may become more productive. 23 More recent work on India points to strong adverse impacts on productivity and on poverty, with poverty rates 3–4 percent higher after 30 years of warming than had warming not occurred. 24 14. Another effect of rising temperature is a likely increase in the volatility of climate outcomes. The evidence suggests that this is likely to increase the vulnerability of the poor in developing countries. 25 Analysis for Tanzania suggests that these increases may be substantial and require adjustments in policies well beyond those thought of conventionally as environmental, such as agricultural trade policies. 26 15. The relationship between trade reform and environmental outcomes is frequently controversial, with some arguing that it raises incomes and improves environmental outcomes while others argue the converse. A book-length study examines six case studies in fragile economies and finds that the outcome depends on specific policies and institutions, and provides lessons for improved policies for sustainable use of natural resources. 27 21 Khandker, Shahidur R., Douglas F. Barnes, Hussain A. Samad, and Nguyen Huu Minh. Forthcoming. “Welfare Impacts of Rural Electrification: Evidence from Vietnam.� Economic Development and Cultural Change. (Based on Policy Research Working Paper 5057, World Bank, Washington, DC.) 22 Khandker, Shahidur R., Douglas F. Barnes, and Hussain A. Samad. 2012. “The Welfare Impacts of Rural Electrification in Bangladesh.� The Energy Journal 33(1): 187–205. 23 Seo, S. N., Robert Mendelsohn, Ariel Dinar, Rashid Hassan, and Pradeep Kurukulasuriya. 2009. “A Ricardian Analysis of the Distribution of Climate Change Impacts on Agriculture across Agro-Ecological Zones in Africa.� Environmental & Resource Economics 43(3): 313–32. 24 Jacoby, Hanan, Mariano Rabassa, and Emmanuel Skoufias. 2011. “Distributional Implications of Climate Change in India.� World Bank Policy Research Working Paper 5623, World Bank, Washington, DC. 25 Ahmed, Syud Amer, Noah Diffenbaugh, and Thomas W. Hertel. 2009. “Climate Volatility Deepens Poverty Vulnerability in Developing Countries. Environmental Research Letters 4: 1–8. 26 Ahmed, Syud Amer, Noah Diffenbaugh, Thomas W. Hertel, and Will Martin. Forthcoming. “Agriculture and Trade Opportunities for Tanzania: Past Volatility and Future Climate Change.� Review of International Economics. 27 Cook, Jonathan A., Owen Cylke, Donald F. Larson, John D. Nash, and Pamela Stedman-Edwards, eds. 2010. Vulnerable Places, Vulnerable People: Trade Liberalization, Rural Poverty and the Environment. World Bank and Edward Elgar; Washington, DC, and North Hampton, MA. 66 Rural-urban migration 16. Rural-urban migration and off-farm work by farmers are frequently major forces for transformation both of agriculture and of the economy in poor countries. A study of nonfarm employment in Bangladesh finds that access to major urban areas matters most for high-return nonfarm activities. 28 By contrast, low-return, nonfarm work is driven purely by local demand. A study of occupational mobility out of agriculture in Nepal and Vietnam finds lower mobility for women, but that mothers’ participation outside agriculture has important impacts on female outmigration rates in Nepal. 29 17. Migration out of rural areas is strongly affected by aspects of land tenure that discourage movement out of agriculture. A study of restrictions on the sale of land in Sri Lanka finds that a 1-percent increase in the share of land subject to sales restrictions reduces rural wages by 6.6 percent at the median distance from an urban center. 30 18. In China, land policies have been shown to effect both migration and on-farm productivity. A history of land reallocations in an area sharply discourages out-migration and lowers productivity by around 30 percent, while provision of land-use certificates encourages migration and raises productivity by roughly the same amount. 31 Price incentives, trade, and vulnerability 19. The levels of producer and consumer prices can have important impacts on well-being given the importance of agriculture as a source of income and in the expenditures of poor people. Historically, developing countries have tended to depress agricultural prices through border measures such as export taxes, while industrial countries have tended to protect their agricultural sectors. A major effort has changed our understanding of the pattern of agricultural distortions, highlighting the near disappearance of agricultural taxation and the emergence of agricultural protection in many developing countries, together with a decline in the rate of protection in developed countries. 32 The data from these studies are readily available and widely used for 28 Deichmann, Uwe, Forhad Shilpi, and Renos Vakis. 2009. “Urban Proximity, Agricultural Potential and Rural Non-farm Employment: Evidence from Bangladesh.� World Development 37(3): 645–60. 29 Emran, M. Shahe, and Forhad Shilpi. 2011. “Intergenerational Occupational Mobility in Rural Economy: Evidence from Nepal and Vietnam.� Journal of Human Resources 46(2): 427–58. 30 Emran, M. Shahe, and Forhad Shilpi. 2010. “General Equilibrium Effects of Land Market Restrictions on Labor Market : Evidence From Wages In Sri Lanka.� World Bank Policy Research Working Paper 5461, World Bank, Washington, DC. 31 Klaus Deininger, Songqing Jin, and Fang Xia. 2012. “Moving Off the Farm: Land Institutions to Facilitate Structural Transformation and Agricultural Productivity Growth in China.� Policy Research Working Paper 5949, World Bank, Washington, DC. 32 Anderson, Kym, and Will Martin, eds. 2009. Distortions to Agricultural Incentives in Asia. Washington, DC: World Bank. Anderson, Kym, and William Masters, eds. 2009. Distortions to Agricultural Incentives in Africa. Washington, DC: World Bank. 67 analytical work. 33 These data also highlight a strong tendency for policy makers in developing countries to counteract changes in world prices through partially offsetting changes in trade distortions. 34 Recent research suggests that greater use of information technology can result in higher returns to farmers, while entry into markets for export crops depends heavily on the availability of outlets at the local level. 35, 36 20. The long-running Doha Agenda includes an ambitious attempt to reform agricultural policies in which agricultural reforms have played a major role. This agenda has become extremely complicated as the negotiators have attempted to address the concerns of over 150 World Trade Organization members. A recent book analyzes the potential impacts of the latest proposals, identifies areas where further progress might be made, and provides lessons for future negotiations. 37 It finds that the proposals would substantially improve agricultural market access for rich and poor countries, despite the economically costly flexibilities included in the agreement. 38 21. The food price spikes of 2008 and 2010 have generated considerable concern about short and long run impacts on poverty and nutritional outcomes. World Bank research has shown that short-term price spikes tend to raise poverty, despite the importance of agriculture in the incomes of the poor. 39 Research on the distributional impacts of trade reforms and price surges also highlights the importance of social safety nets in dealing with the poverty impacts of price changes. 40 Recent research provides a rationale for the observed tendency of developing countries to use varying trade distortions—preferably accompanied by active storage policies— in an attempt to partially stabilize domestic prices. 41 One concern with the widespread use of Anderson, Kym, ed. 2009. Distortions to Agricultural Incentives: A Global Perspective, 1955-2007. Washington, DC: World Bank. 33 Anderson, Kym, John Cockburn, and Will Martin, eds. 2010. Agricultural Price Distortions, Inequality, and Poverty. Washington, DC: World Bank. 34 Anderson, Kym, ed. 2010. The Political Economy of Agricultural Price Distortions. New York: Cambridge University Press. 35 Goyal, Aparajita. 2010. “Information, Direct Access to Farmers, and Rural Market Performance in Central India.� American Economic Journal: Applied Economics 2(3): 22–45. 36 Balat, Jorge, Irene Brambilla, and Guido Porto. 2009. “Realizing the Gains from Trade: Export Crops, Marketing Costs, and Poverty.� Journal of International Economics 78(1): 21–31. 37 Martin, Will, and Aaditya Mattoo, eds. 2011. Unfinished Business? The WTO’s Doha Agenda. London: CEPR and Washington, DC: World Bank. 38 Jean, Sébastien, David Laborde, and Will Martin. 2010. “Formulas and Flexibility in Trade Negotiations: Sensitive Agricultural Products in the World Trade Organization’s Doha Agenda.� World Bank Economic Review 24(3): 500–19. 39 Ivanic, Maros, Will Martin, and Hassan Zaman. 2011. “Estimating the Short-Run Poverty Impacts of the 2010–11 Surge in Food Prices.� Policy Research Working Paper 5633, World Bank, Washington, DC. (World Bank Food Price Watch, April 2011.) 40 Coady, David P., Paul A. Dorosh, and Bart Minten. 2009. “Evaluating Alternative Policy Responses to Higher World Food Prices: The Case of Increasing Rice Prices in Madagascar.� American Journal of Agricultural Economics 91(3): 711–22. 41 Gouel, Christophe, and Sébastien Jean. 2012. “Optimal Food Price Stabilization in a Small Open Developing Country.� Policy Research Working Paper 5943, World Bank, Washington, DC. 68 such price-insulating trade distortions is that they can increase the volatility of world prices, reducing the effectiveness of countries’ attempt to stabilize domestic prices. 42 22. Prepared by Will Martin, Research Manager, Development Research Group, Development Economics Vice Presidency. A.2 Environment, Energy, and Sustainable Development Research Program 23. World Bank research on energy, environment, water, and climate change provides information and insight on integrating environmentally sustainability with economic efficiency in advancing growth and poverty reduction. Issues concerning urban-regional development and infrastructure cut across all these topic areas. Themes 24. Energy is both a key factor in economic development and a central focus in addressing environmental problems and climate change. Energy research examines the economic and environmental implications of alternative energy resources, increased access to energy services, and improved efficiency in public utility regulation. 25. Research on environmental, water, and other natural resource problems in developing countries examines policies for reducing pollution and congestion in urban transportation, the economic value of environmental and natural resources, and nontraditional policies for their protection. 26. Climate change poses significant risks through its effects on temperature, precipitation, sea level, and natural disasters. Climate change research examines the implications of these risks for a wide range of investment decisions, as well as policies for reducing climate change risks and mitigating the greenhouse gases that contribute to climate change. 27. Development is accompanied by significant spatial transformations—changes in the distribution of people and economic activities. Adequate infrastructure is needed in turn to stimulate and sustain development. Research on these topics focuses particularly on urban density, access to markets, and impacts of inter-regional transport improvements on trade flows and economic performance, as well as on the importance of energy availability. 42 Martin, Will, and Kym Anderson. 2011. “Export Restrictions and Price Insulation during Commodity Price Booms.� Policy Research Working Paper 5645, World Bank, Washington, DC. Anderson, Kym, and Signe Nelgen. 2010. “Trade Barrier Volatility and Domestic Price Stabilization: Evidence from Agriculture.� Policy Research Working Paper 5511, World Bank, Washington, DC. 69 Highlights Alternative energy resources present a complex mix of pros and cons 28. Even with greatly improved energy efficiency, demand for different forms of energy will increase significantly as incomes in developing countries rise. The challenge is to satisfy growth in demand affordably and sustainably. 43 Coal is plentiful and cheap, and its use can be easily scaled-up, but expanded coal use with current technologies intensifies already strong concerns about global warming and local air pollution. Liquid biofuels made from plant matter can be a low-emissions resource, but current food-crop based sources require subsidies for large-scale production, and may not decrease greenhouse emissions substantially—especially if expanded supplies increase deforestation. By competing with food supplies, a global expansion of biofuels could cause important economic losses in some regions, though other regions would gain economically from being biofuels sellers. 44 Better biofuel production technologies exist, but it will be some years before they are technologically and economically feasible on a large scale. 45 29. Wind and solar energy for electricity generation have advanced significantly over the past two decades. However, the intermittency of these sources and the need to connect together an enormous number of far-flung, small resources remain significant technical and economic challenges to deploying them at a scale that would replace large quantities of fossil fuels. 46,47 Expansion of nuclear power faces the opposite set of challenges. It can produce a steady supply of electricity at a large scale, but it currently faces an unknown future given concerns about safety and proliferation, as well as a high and still-uncertain initial capital cost. 48 Ultimately, all of these and other energy resources will be needed to successfully meet growth in global demand. 43 Kessides, Ioannis N., and David C. Wade. 2011. “Towards a Sustainable Global Energy Supply Infrastructure: Net Energy Balance and Density Considerations.� Energy Policy 39(9): 5322–34. 44 Timilsina, Govinda R., John C. Beghin, Dominique van der Mensbrugghe, and Simon Mevel. Forthcoming. “The Impacts of Biofuel Targets on Land-Use Change and Food Supply: A Global CGE Assessment.� Agriculture Economics. (Based on Policy Research Working Paper 5513, World Bank, Washington, DC.) 45 Carriquiry, Miguel A., Xiaodong Du, and Govinda R Timilsina. 2011. “Second Generation Biofuels: Economics and Policies.� Energy Policy 39(7): 4222–34. 46 Timilsina, Govinda R., Lado Kurdgelashvili, and Patrick A. Narbel. 2012. “Solar Energy: Markets, Economics and Policies.� Renewable and Sustainable Energy Reviews 16(1): 449-465. 47 van Kooten, G. Cornelis, and Govinda R. Timilsina. 2009. “Wind Power Development: Economics and Policies.� Policy Research Working Paper 4868, World Bank, Washington, DC. 48 Adamantiades, Achilles, and Ioannis N. Kessides. 2009. “Nuclear Power for Sustainable Development: Current Status and Future Prospects.� Energy Policy 37(12): 5149–66. 70 No single solution for increasing access to affordable and clean energy 30. Lack of access to reliable, affordable, and safe energy—“energy poverty� —is a key barrier to economic development in both rural and peri-urban areas. 49 The challenge is how this is to be accomplished. In many cases, rural households can obtain the limited quantities of energy they need through small-scale, decentralized sources, often using locally available renewable energy sources. Even in Africa, however, the trend toward urbanization and the current cost premium paid for small-scale energy can tip the balance toward extending the electricity grid into areas not too far away. 50 Aside from choice of technology, however, the necessary institutional measures for monitoring and charging for electricity use must be put in place in order to maintain financial sustainability. A study of illegal electricity connections in the favelas of Belo Horizonte, Brazil found that electricity theft was motivated not just by low incomes, but also by poor quality of service and equipment. 51 Such a situation can create a vicious circle in which poor service undermines the willingness of costumers to pay for service, which exacerbates funding shortages for upgrading service. 52 Several tools needed to solve environmental and congestions problems with urban transportation 31. As cities grow throughout the developing world, so do transport-related problems— traffic congestion, local air pollution, and increased greenhouse gas emissions. Continued road investments lock in urban sprawl and private reliance on private vehicles, making it harder to reduce time wasted in traffic congestion and pollution from automobile emissions. Part of the solution is investments in public transportation that connect urban centers with each other and with surrounding areas. 53 However, policies to limit transportation services demands also are needed, since the price of fuels does not reflect the cost of the resulting pollution and with a very few exceptions, there is no direct pricing of road access that could reduce congestion. Even without road pricing, there could be significant overall benefits from raising fuel taxes to reflect pollution and congestion “externalities.� However, a study of externality pricing in México City indicates that reaping these gains would require significant increases in fuel prices and total driving cost. 54 The full social costs of road travel needs to be considered in evaluating alternative 49 Barnes, Douglas F., Shahid R. Khandker, and Hussain A. Samad. 2011. “Energy Poverty in Rural Bangladesh.� Energy Policy 39(2): 894–904. 50 Deichmann, Uwe, Craig Meisner, Siobhan Murray, and David Wheeler. 2011. “The Economics of Renewable Energy Expansion in Rural Sub-Saharan Africa.� Energy Policy 39(1: 215-227. 51 Mimmi, Luisa M., and Sencer Ecer. 2010. “An Econometric Study of Illegal Electricity Connections in the Urban Favelas of Belo Horizonte, Brazil.� Energy Policy 38(9): 5081-5097. 52 Strand, Jon. 2012. “Low-Level versus High-Level Equilibrium in Public Utility Services.� Journal of Public Economics 96: 163–72. 53 Anas, Alex, and Govinda Timilsina. 2009. “Lock-in Effects of Road Expansion on CO2 Emissions: Results from a Core-Periphery Model of Beijing.� Policy Research Working Paper 5017, World Bank, Washington, DC. 54 Parry, Ian W. H., and Govinda R. Timilsina. 2010. “How Should Passenger Travel in Mexico City Be Priced?� Journal of Urban Economics 682): 167–82. 71 transportation infrastructure investments and designing measures to finance the necessary infrastructure. People value water access and cleanliness—but how to provide it in practice? 32. There is significant evidence that poor people put a relatively high economic value on improved access to water and on the cleanliness of their water supply. 55,56 Yet water supplies are often inadequate and nowhere near potable. Evidence shows that people would be willing to pay more for improved water. 57 In practice, however, increasing water tariffs can have a mix of desired and unwanted effects: water quality and availability may improve, reducing adverse health effects of contaminated water use, but the necessary cost increases could impose a significant burden on the poor and indirectly limit access to water supplies in rural areas. 58 There is also need for social mechanisms to induce cooperation in limiting pollution of community water supplies. 59 Alternatives to traditional regulation can extend the reach of environmental protection 33. In many parts of the developing world, the mechanisms for limiting degradation of environmental and natural resources—pollution standards, land use controls, taxes and fines— are not readily applicable because the institutions for their use are not yet adequately developed. However, some alternative approaches show promise. A “Payment for Environmental Services� (PES) system provides financial incentives for those in charge of sensitive resources to reduce their degradation and promote more sustainable use of them. Financing for such services can come from direct user fees, indirect taxes on environmentally damaging goods, or general revenues. Provided performance can be adequately monitored and payments are sufficiently large to motivate changes in behavior, the system can achieve its environmental goals. 60 34. Information disclosure programs and environmental performance ratings also can provide an alternative to difficult-to-implement pollution regulation. In China’s Jiangsu Province, the 55 Nauges, Celine, Jon Strand, and Ian Walker. 2009. “The Value of Water Connections in Central American Cities: A Revealed Preference Study.� Environment and Development Economics 14(3): 349–70. 56 Wang, Hua, Yuyan Shi, Yoonhee Kim, and Takuya Kamata. 2011. “Valuing Water Quality Improvement in China: A Case Study of Lake Puzhehei in Yunnan Province.� Policy Research Working Paper 5766, World Bank, Washington, DC. 57 Wang, Hua, Jian Xie, and Honglin Li. 2010. “Water Pricing with Household Surveys: A Study of Acceptability and Willingness to Pay in Chongqing, China.� China Economic Review 21(1): 136–49. 58 Barrera-Osorio, Felipe, and Mauricio Olivera. 2009. “Does Society Win or Lose as a Result of Privatization? The Case of Water Sector Privatization in Colombia.� Economica 76(304): 649–74. 59 Kremer, Michael, Jessica Leino, Edward Miguel, and Alix Peterson Zwane. 2011. “Spring Cleaning: Rural Water Impacts, Valuation, and Property Rights Institutions.� Quarterly Journal of Economics 126(1): 145–205. 60 Pagiola, Stefano, Ana R. Rios, and Agustin Arcenas. 2010. “Poor Household Participation in Payments for Environmental Services: Lessons from the Silvopastoral Project in Quindío, Colombia.� Environmental and Resource Economics 47(3): 371–94. 72 “Green Watch� performance rating and disclosure program has led to greater pollution reduction among rated companies than non-covered firms, after adjusting for differences among firm characteristics. 61 Environmental assets and other intangible wealth are important factors in economic development 35. One of the puzzles of national wealth accounting is that the value of tangible assets— fixed capital and land—is a smaller multiple of national income than expected given realistic rates of return. New and more comprehensive national wealth measures indicate that while intangible wealth, including a number of environmental assets, is not measured in national accounts, it represents 60 to 80 percent of comprehensive wealth across most countries. Increasing environmental assets thus is one key contributor to the growth and development of nations. 62,63 Natural hazards exacerbated by climate change will require a range of responses 36. Developing countries cope with several types of natural disasters that are likely to be exacerbated by climate change, including storm surges that overwhelm coastal areas as well as other impacts of tropical storms. 64,65 Increasing and more dense urbanization in pursuit of greater economic opportunity will greatly increase the number of city inhabitants exposed to high risks of tropical cyclones, especially with inadequate risk prevention and mitigation measures as simple as unblocking drains. 66 However, there is also evidence that the risk of dying from flooding decreases with income, so higher incomes in urban areas would lessen risks. 67 Effectively responding to growing risks from natural disasters with climate change necessitates a mix of larger-scale and smaller-scale measures over time. In Bangladesh, for example, where much of the country is less than 5 meters above sea level, major improvements in dikes can be combined with reforestation of coastlines, strengthening of private homes (with provision for 61 Jin, Yanhong, Hua Wang, and David Wheeler. 2010. “Environmental Performance Rating and Disclosure: An Empirical Investigation of China’s Green Watch Program.� Policy Research Working Paper 5420, World Bank, Washington, DC. 62 Ferreira, Susana, and Kirk E. Hamilton. 2010. “Comprehensive Wealth, Intangible Capital, and Development.� Policy Research Working Paper 5452, World Bank, Washington, DC. 63 World Bank. 2011. The Changing Wealth of Nations: Measuring Sustainable Development in the New Millennium. Washington, DC: World Bank. 64 Dasgupta, Susmita, Benoit Laplante, Siobhan Murray, and David Wheeler. 2011. “Exposure of Developing Countries to Sea-Level Rise and Storm Surges.� Climatic Change106: 567–79. 65 Mendelsohn, Robert, Kerry Emanuel, and Shun Chonabayashi. 2011. “The Impact of Climate Change on Global Tropical Storm Damages.� Policy Research Working Paper 5562, World Bank, Washington, DC. 66 Lall, Somik V., and Uwe Deichmann. Forthcoming. “Density and Disasters: Economics of Urban Hazard Risk.� World Bank Research Observer (doi: 10.1093/wbro/lkr006, July 7). 67 Ferreira, Susana, Kirk Hamilton, and Jeffrey R. Vincent. 2011. “Nature, Socioeconomics and Adaptation to Natural Disasters: New Evidence from Floods.� Policy Research Working Paper 5725, World Bank, Washington, DC. 73 also sheltering livestock), and improvement in early warning systems. 68 Agreements among countries for use of shared water bodies will have to adjust to reflect changes in precipitation runoff. 69 Farmers in Africa and elsewhere have several potential options for adapting to climate change 37. Agricultural crops and grazing lands comprise 40 percent of global land use, and in many developing countries agriculture provides employment and livelihood for three-quarters of the population. More than 15 years of Bank research on the impact of, and long-term adaptation to, climate change in agriculture under a variety of conditions has been synthesized in a recent book spanning more than 22 countries across four continents. 70 Comparison of long-term adjustments across countries is used to examine the impact of climate change on agriculture and livestock, as well as to quantify how farmers adapt to climate change. Because cropping is more sensitive than grazing to climate change, and rain-fed cropland is generally more sensitive than irrigated cropland, farmers may be able to adjust to climate change by switching crops and livestock species, increasing irrigation, and moving between livestock and crops. However, research shows that future farm incomes in Africa are very climate sensitive, and will be severely threatened in the event of extreme climate change scenarios. 71 Adaptation policies and measures also must be location-specific. Public policies to increase resilience through investments in knowledge diffusion and infrastructure can increase productivity across a range of future climate scenarios. International cooperation to reduce greenhouse gases remains limited 38. Climate change is a global problem that can only be met with global cooperation to reduce greenhouse gas emissions. Cooperation between developed and developing countries for emissions mitigation has broadened over the past decade, but it remains limited in scope and impact. The Clean Development Mechanism (CDM), a voluntary mechanism for developed countries to help fund projects in developing countries, has grown in size, but there are lingering questions about how well it operates and about how many projects are “additional� to activities 68 Dasgupta, Susmita, Mainul Huq, Zahirul Huq Khan, Manjur Murshed Zahid Ahmed, Nandan Mukherjee, Malik Fida Khan, and Kiran Pandey. 2010. “Vulnerability of Bangladesh to Cyclones in a Changing Climate: Potential Damages and Adaptation Cost.� Policy Research Working Paper 5280, World Bank, Washington, DC. 69 Dinar, Ariel, Brian Blankespoor, Shlomi Dinar, and Pradeep Kurukulasuriya. 2010. “Does Precipitation and Runoff Variability Affect Treaty Cooperation between States Sharing International Bilateral Rivers?� Ecological Economics 69(12): 2568–81. 70 Mendelsohn, Robert, and Ariel Dinar. 2009. Climate Change and Agriculture: An Economic Analysis of Global Impacts, Adaptation and Distributional Effects. Washington, DC: World Bank, and North Hampton, Mass.: Edward Elgar. 71 Seo, S. N., Robert Mendelsohn, Ariel Dinar, Rashid Hassan, and Pradeep Kurukulasuriya. 2009. “A Ricardian Analysis of the Distribution of Climate Change Impacts on Agriculture across Agro-ecological Zones in Africa.� Environmental and Resource Economics 43(3): 313–32. 74 that would have been undertaken anyway. 72,73 Unless stronger international agreements are reached to actually reduce greenhouse gas emissions, it will be difficult for broader “market mechanisms� like the CDM (or higher energy taxes to limit emissions) to be instituted. More piecemeal interim measures may have to implemented, instead. 39. One possibility is putting internationally coordinated taxes on aviation and maritime fuels, which so far have been exempted from such measures under a variety of international treaties. Such taxes could have a small but noticeable effect on global emissions while raising billions of dollars each year that could be used for “climate smart� purposes. Even if lower- income developing countries have tax revenues returned to them in the interest of international equity, obtaining international agreements to change multiple treaties ruling out such taxes would be a formidable task. 74 Ultimately such sector-based measures could be aggregated into an architecture for emissions mitigation based on a variety of approaches tailored to differing circumstances in the various sectors. For example, agreements to limit the emissions intensity of countries’ electricity sectors on the one hand, with agreed changes in technical standards and practices to improve building energy efficiencies. 75 “Connectedness� of infrastructure and institutions enhances regional growth in Africa 40. The benefits of expanding trade within Sub-Saharan Africa by upgrading and maintaining the trunk road network connecting capitals and the largest cities is estimated to be in the tens of billions of dollars each year—well in excess of costs—with additional benefits accruing from greater within-country trade, increased access to global markets, and considerable employment effects. 76 Similarly, harmonization of legal, regulatory and institutional frameworks along with development of regional-scale “backbone� infrastructure for “network� industries like telecommunications or electricity can lead to reduced cross-border transaction costs, greater economies of scale, and the potential for increased competition, as illustrated by examining regionalized telecommunications policy in West Africa. 77 However, a country’s ability to benefit 72 Rahman, Shaikh M., Ariel Dinar, and Donald F. Larson. 2010. “Will the Clean Development Mechanism Mobilize Anticipated Levels of Mitigation?� Policy Research Working Paper Series 5239, World Bank, Washington, DC. 73 Strand, Jon, and Knut Einar Rosendahl. 2011. “Carbon Leakage from the Clean Development Mechanism.� Energy Journal 32(4): 27-50. 74 Keen, Michael, Ian Parry, and Jon Strand. 2012. “Market-Based Instruments for International Aviation and Shipping as a Source of Climate Finance.� Policy Research Working Paper 5950, World Bank, Washington, DC. 75 Barrett, Scott, and Michael Toman. 2010. “Contrasting Future Paths for an Evolving Global Climate Regime.� Global Policy 1(1): 64-74. 76 Buys, Piet, Uwe Deichmann, and David Wheeler. 2010. “Road Network Upgrading and Overland Trade Expansion in Sub-Saharan Africa.� Journal of African Economics 19(3): 399–432. 77 Kessides, Ioannis N., Roger G. Noll, and Nancy C. Benjamin. 2009. “Regionalizing Telecommunications Reform in West Africa.� Policy Research Working Paper 5126, World Bank, Washington, DC. 75 from such spillover effects also depends on broader institutional factors, such as participation in formal trade agreements and how well agreements function in practice. 78 China’s expressway construction supports growth, but regional disparities remain 41. China’s ambitious program of expressway network expansion over the past two decades is estimated to have raised national income in 2007 approximately 6 percent above what it otherwise would have been. However, the expressway network appears to have reinforced existing patterns of spatial income inequality, although the results vary across the country. While increased disparities may stimulate further migration to coastal growth centers in the short term, there is evidence from more advanced developed countries that the disparities will decline in the longer term—especially if infrastructure investments that promote urbanization and economic concentration are accompanied by policies that ensure access to health, education, and other basic services in rural and lagging areas. 79 Prepared by Michael A. Toman, Research Manager, Development Research Group, Development Economics Vice Presidency. A.3 Finance and Private Sector Development Research Program 42. Finance and Private Sector Development research focuses on understanding the role of the financial and private sectors in promoting economic development and reducing poverty, and identifying policies to improve their effectiveness. Finance research is organized around access to financial services and risk management and stability. Private sector research focuses on determinants of firm entry and performance to better understand the microeconomics of the growth process. Themes 43. The research on access to finance—which is important in promoting growth and alleviating poverty—includes documenting and benchmarking access to financial services by small firms and poor households, and identifying underserved groups and analyzing barriers to building more inclusive financial systems. Evaluative research studies the channels through 78 Roberts, Mark, and Uwe Deichmann. 2009. “International Growth Spillovers, Geography and Infrastructure.� Policy Research Working 5153, World Bank, Washington, DC. 79 Roberts, Mark, Uwe Deichmann, Bernard Fingleton, and Tuo Shi. 2010. “On the Road to Prosperity? The Economic Geography of China’s National Expressway Network.� Policy Research Working Paper 5479, World Bank, Washington, DC. 76 which access to finance can contribute to growth through promoting entrepreneurship, innovation, and the process of technology adoption. 44. Research on the private sector focuses on firm dynamics—changes in the composition of the private sector and entry and exit over time—and performance, and how this affects firm productivity and growth. Special areas of focus include research on the determinants and consequences of informality, innovation, and governance as well as the impact of the business environment and its reforms. Highlights Have the rules of finance changed after the crisis? 45. The global financial crisis which started in 2008 challenged conventional thinking in areas of finance and private sector development, and continues to reveal key knowledge gaps. First, how do we ensure that financial systems are growth-promoting and inclusive, yet stable? What are the key tradeoffs? What roles do financial liberalization, regulation and supervision, and financial education play in this process? Second, are there optimal financial structures at different stages of the development process? What does this imply for the role of banking versus market development, banking structure, and size distribution? And finally, which policy interventions enhance the role of the financial and private sector in promoting development? 80 46. Research on the causes of the crisis attributes a significant role to faulty microeconomic incentives. 81 It has also investigated the implications of the crisis for macroeconomic and financial policies, emphasizing that the challenge of financial sector policies is to align private incentives with public interest without taxing or subsidizing private risk-taking. The analysis suggests that public ownership or regulation that is too aggressive would simply hamper financial development and growth. But clearly striking this balance is becoming increasingly complex in an ever more integrated and globalized financial system. 82,83 47. Investigating the financial system trends in the boom period leading up to the global crisis shows that lower margins from traditional lines of business and the search for higher returns were only possible through high risk-taking, especially in high-income countries. 84 Analyzing bank activity and funding strategies during this period also shows that 80 Demirgüç-Kunt, Asli, Douglas D. Evanoff and George G. Kaufman, eds. 2011. The International Financial Crisis: Have the Rules of Finance Changed? New Jersey: World Scientific Publishing Company. 81 Caprio, Gerard, Asli Demirgüç-Kunt, and Edward Kane. 2010. “The 2007 Meltdown in Structured Securitization: Searching for Lessons, not Scapegoats.� World Bank Research Observer 25(1): 125–55. 82 Demirgüç-Kunt, Asli, and Luis Servén. 2010. “Are All Sacred Cows Dead? Implications of the Financial Crisis for Macro and Financial Policies.� World Bank Research Observer 25(1): 91–124. 83 de la Torre, Augusto, and Alain Ize. 2010. “Regulatory Reform: Integrating Paradigms.� International Finance 13(1): 109-139. 84 Beck, Thorsten, Asli Demirgüç-Kunt, and Ross Levine. 2010. “Financial Institutions and Markets across Countries and over Time.� World Bank Economic Review 24(1): 77–92. 77 overall, banking strategies that rely prominently on generating noninterest income or attracting nondeposit funding are very risky, consistent with the demise of the U.S. investment banking sector. 85 The recent financial crisis can also teach us lessons on bank capital regulation. During the crisis, higher capital resulted in better stock performance, mostly for larger banks and less well-capitalized banks, suggesting that calls to strengthen capital requirements is broadly appropriate. The relationship between stock returns and capital is stronger when capital is measured by the leverage ratio rather than the risk-adjusted capital ratio, indicating how difficult it is to properly measure risk exposure for large and complex financial organizations. Finally, higher quality forms of capital, such as Tier 1 capital and tangible common equity are likely to provide more effective protection. 86 48. Our research on re-examining the role of the state in the financial sector, particularly on regulation and supervision, ownership of financial institutions, competition policy and strengthening financial infrastructure is feeding into a new Bank publication, Global Financial Development Report, which is expected to position the Bank Group more prominently in policy discussions in the area of financial development. Financial services remain out of reach for many in developing countries 49. Data show that more than half of people in developing countries face financial exclusion due to high barriers in accessing financial services. A special issue of the World Bank Economic Review documents barriers to access, and investigates the impact of access to finance on households and firms. 87 Theory and empirical evidence point to the critical role that improved access to finance has in promoting growth and reducing income inequality. 88 50. Papers in the symposium developed first estimates of financial outreach across countries, assembled evidence on the impact of finance on firm performance, and presented initial results 85 Demirgüç-Kunt, Asli, and Harry Huizinga. 2010. “Bank Activity and Funding Strategies: The Impact on Risks and Returns.� Journal of Financial Economics 98(3): 626–50. 86 Demirgüç-Kunt, Asli, Enrica Detragiache, and Ouarda Merrouche. 2010. “Bank Capital: Lessons from the Financial Crisis.� Policy Research Working Paper 5473, World Bank, Washington, DC. 87 Beck, Thorsten, and Asli Demirgüç-Kunt, ed. 2008. “Special Issue: Access to Finance.� World Bank Economic Review 22(3, October). Selected articles from this issue include: Ayyagari, Meghana, Asli Demirgüç-Kunt, and Vojislav Maksimovic. 2008. “How Important Are Financing Constraints? The Role of Finance in the Business Environment.� World Bank Economic Review 22(3): 483–516. Beck, Thorsten, and Asli Demirgüç-Kunt. 2008. “Access to Finance: An Unfinished Agenda.� World Bank Economic Review 22(3): 383–96. Beck, Thorsten, Asli Demirgüç-Kunt, and María Soledad Martínez Pería. 2008. “Banking Services for Everyone? Barriers to Bank Access and Use around the World.� World Bank Economic Review 22(3): 397–430. McKenzie, David, and Christopher Woodruff. 2008. “Experimental Evidence on Returns to Capital and Access to Finance in Mexico.� World Bank Economic Review 22(3): 457–82. Gine, Xavier, Robert Townsend, and James Vickery. 2008 “Patterns of Rainfall Insurance Participation in Rural India.� World Bank Economic Review 22(3): 539–66. 88 Demirgüç-Kunt, Asli, and Ross Levine. 2009. “Finance and Inequality: Theory and Evidence.� Annual Review of Financial Economics 1: 287–318. 78 on techniques and products to reach out to micro-borrowers and savers, including insurance. These papers were some of the background research for the Policy Research Report Finance for All? Policies and Pitfalls in Expanding Access, which was disseminated widely in 2008 in many of our client countries. 89 One of the important messages of the report is on lack of data on financial inclusion. In 2010, the Bill and Melinda Gates Foundation gave a grant to the Development Research Group to construct a new public database on “Global Financial Inclusion Indicators,� to measure how the world’s poor, women, and other disadvantaged groups save, borrow and make payments. The goal of the project is to measure financial inclusion in a consistent manner over a broad range of countries and over time, and create a public dataset that can be used to motivate, formulate, and track global policy and progress to improve access to financial services, and hence contribute to the G-20 agenda on financial inclusion. 51. Microfinance meets the market, but with high transaction costs. Microfinance institutions have proved the possibility of providing reliable banking services to poor customers. Their second aim is to do so in a commercially viable way. Research analyzed the tensions and opportunities of microfinance as it embraces the market, drawing on a dataset that includes 346 of the world’s leading microfinance institutions and covers nearly 18 million active borrowers. 52. The data show high rates of loan repayment, but also suggest that profit-maximizing investors would have limited interest in most of the institutions focusing on the poorest customers and women. Those institutions, as a group, charge their customers the highest fees in the sample but also face particularly high transactions costs, in part due to small transaction sizes. Innovations to overcome well-known information problems in financial markets were a triumph, but further innovation is needed to overcome the challenges of high costs. 90 Regulatory supervision is also an important issue affecting microfinance profitability and outreach. 91 53. Do large and foreign banks serve small and medium businesses? Conventional wisdom is skeptical. But the data show that all types of banks are catering to SMEs and large, multiple- service banks have a comparative advantage in offering a wide range of products and services on a large scale, through the use of new technologies, business models, and risk management systems. Moreover, these patterns have not been derailed by the global financial crisis. 92, 93 54. Role of financial literacy in raising demand for formal financial services in emerging markets.Financial literacy campaigns are increasingly heralded as the best way to increase 89 World Bank. 2008. Finance for All? Policies and Pitfalls in Expanding Access. A World Bank Policy Research Report. Washington, DC: World Bank. Download at: http://go.worldbank.org/J44HH7AGY0. PRR website: http://econ.worldbank.org/prr. 90 Cull, Robert J., Asli Demirgüç-Kunt, and Jonathan Morduch. 2009. “Microfinance Meets the Market.� Journal of Economic Perspectives 23(1): 167–92. 91 Cull, Robert J., Asli Demirgüç-Kunt, and Jonathan Morduch. 2011. “Does Regulatory Supervision Curtail Microfinance Profitability and Outreach?� World Development 39(6): 949–65. 92 de la Torre, Augusto, María Soledad Martínez Pería, and Sergio L. Schmukler. 2010. “Bank Involvement with SMEs: Beyond Relationship Lending.� Journal of Banking and Finance 34(9): 2280-93. 93 Beck, Thorsten, Asli Demirgüç-Kunt, and María Soledad Martínez Pería. 2011. “Bank Financing for SMEs: Evidence across Countries and Bank-Ownership Types.� Journal of Financial Services Research 39(1): 35-54. 79 demand for formal financial services in emerging markets. But do they work and for whom? One view argues that limited financial literacy stifles demand. A second view argues that demand is rationally low, because formal financial services are expensive and of relatively low value to the poor. 55. Original surveys and a field experiment in India and Indonesia show that financial literacy is a powerful predictor of demand for financial services. The relative importance of literacy and price was tested in a field experiment that offered randomly selected unbanked households financial literacy education crossed with a small incentive (US $3–$14) to open a bank savings account. The financial literacy program had no effect on the likelihood of opening a bank savings account in the full sample, but showed modest effects for uneducated and financially illiterate households. In contrast, small subsidy payments had a large effect on the likelihood of opening a savings account. These payments proved to be more than two times more cost-effective than the financial literacy training. 94 56. Technology can help improve financial access: Biometric identification can increase loan repayment rates. In countries that lack a unique identification system, identity fraud—the use of someone else’s identity or a fictitious one—is a common means of gaining access to credit or other services. Lenders tell anecdotes of past borrowers purposefully defaulting and trying to obtain a fresh loan from the same or another institution. As a result, lenders have restricted the supply of credit. In a field experiment, randomly selected loan applicants had a fingerprint collected as part of the loan application. Such biometric technology can make the threat of future credit denial credible because it makes it easier for the bank to withhold new loans from past defaulters, and to reward responsible past borrowers with increased credit. The results show that fingerprinted borrowers have substantially higher repayment rates than other borrowers. This is particularly true among borrowers who have high default risk. The increase in repayment rates is due to both a reduction in adverse selection (fingerprinted high-risk borrowers took out smaller loans) and to lower moral hazard (fingerprinted borrowers diverted less of the loan from its intended purpose). A cost-benefit analysis of the pilot experiment suggests that the benefits from improved repayment greatly outweigh the costs of equipment and fingerprint collection. 95 What is needed to promote firm growth and create jobs? 57. The recent financial crisis and the resulting high unemployment have sparked a renewed interest in government policies to create jobs and where job creation occurs. 96 In Mexico, business environment reforms as well expanding access to finance have created jobs by fostering 94 Cole, Shawn, Thomas Sampson, and Bilal Zia. Forthcoming. “Money or Knowledge? What Drives Demand for Financial Services in Emerging Markets?� Journal of Finance. 95 Giné, Xavier, Jessica Goldberg, and Dean Yang. Forthcoming. “Identification Strategy: A Field Experiment on Dynamic Incentives in Rural Credit Markets.� American Economic Review. (Based on Policy Research Working Paper 5438, World Bank, Washington, DC.) 96 Ayyagari, Meghana, Asli Demirgüç-Kunt, and Vojislav Maksimovic. 2011. “Small vs. Young Firms across the World: Contribution to Employment, Job Creation, and Growth.� Policy Research Working Paper 5631. World Bank, Washington, DC. 80 entrepreneurship. Simplified business registration procedures led to an increase in the number of registered firms and raised employment by 2.8 percent. 97 Research also shows that larger and broader reforms are better for stimulating new firm registrations. Countries that start out with high registration costs need larger reforms to induce a significant number of new registrations. These results suggest that small incremental reforms are not as effective as larger broader reforms in boosting the private sector. 98 Opening banks for low-income individuals increased the number of informal businesses, which was also accompanied by a rise in employment, by 1.4 percent. 99 58. Recent labor market regulations on Chinese firms affected firm performance, labor wages, and job creation, leading to labor downsizing with negative short-term effects on productivity, but smaller productivity losses for smaller private firms compared to state-owned firms (though the reverse is true for wage losses), indicating a stronger profit incentive for private firms. 100 Using new data—the World Bank Group Entrepreneurship Survey—and Enterprise Surveys research has also examined a broader set of policies and interventions that could promote entrepreneurial activity and create jobs. 101, 102,103 59. Microenterprises are an important source of employment in most developing countries. Yet few of the self-employed grow beyond subsistence level. A large share of employment in developing countries takes place in small and informal firms. The rapid expansion of microfinance is based on the belief that these firms have productive investment opportunities. Measuring the return to capital, however, is complicated by unobserved factors such as entrepreneurial ability and demand shocks. Researchers used a randomized experiment to measure the return to capital for microenterprises in Sri Lanka. They provided cash and equipment grants to randomly selected small firms and measured the consequent increase in profits from this extra capital. They find the average real return to be 4–5 percent per month, substantially higher than the market interest rate. Returns are highest for poor, high-ability 97 Bruhn, Miriam. 2011. “License to Sell: The Effect of Business Registration Reform on Entrepreneurial Activity in Mexico.� Review of Economics and Statistics 93(1): 382–86. 98 Klapper, Leora, and Inessa Love. 2010. “The Impact of Business Environment Reforms on New Firm Registrations.� Policy Research Working Paper 5493, World Bank, Washington, DC. 99 Bruhn, Miriam, and Inessa Love. 2009. “The Economic Impact of Banking the Unbanked: Evidence from Mexico.� Policy Research Working Paper 4981, World Bank, Washington, DC. 100 Dong, Xiao-Yuan, and L. Colin Xu. 2009. “Labor Restructuring in China’s Industrial Sector: Toward a Functioning Urban Labor Market.� Journal of Comparative Economics 37(2): 287–305. 101 Klapper, Leora, Raphael Amit, and Mauro F. Guillén. Forthcoming. “Entrepreneurship and Firm Formation across Countries.� In International Differences in Entrepreneurship, ed. Joshua Lerner and Antoinette Schoar. Cambridge, Mass.: National Bureau of Economic Research. 102 Ayyagari, Meghana, Asli Demirgüç-Kunt, and Vojislav Maksimovic. Forthcoming. “Firm Innovation in Emerging Markets: The Role of Finance, Governance, and Competition.� Journal of Financial and Quantitative Analysis. (Based on Policy Research Working Paper 4157, World Bank, Washington, DC.) 103 Djankov, Simeon. 2009. “The Regulation of Entry: A Survey.� World Bank Research Observer 24(2): 183-203. Djankov, Simeon, Tim Ganser, Caralee McLiesh, Rita Ramalho, and Andrei Shleifer. 2010. “The Effect of Corporate Taxes on Investment and Entrepreneurship.� American Economic Journal: Macroeconomics 2(3): 31-64. 81 owners. 104 The returns to capital from the grants vary dramatically by gender—large male- owned enterprises experienced higher capital returns then female-owned microenterprises. 105 One argument to explain this outcome is that female-owned enterprises are constrained by the sectors the owners operate in, and by household bargaining issues, suggesting job creation efforts also need to focus on the types of jobs created, not just whether there is a job. 60. Another constraint to the growth of small enterprises is often thought to be informality. Research in Bolivia shows large gains to small enterprises from formalizing, but only for those firms that are constrained by information or distance from registering their businesses. Sole proprietorships appear to not benefit from formalizing. 106 61. Turning to productivity improvements in small and medium enterprises, research shows poor management skills can hamper firm productivity. Free management consulting was provided to a random sample of large Indian textile firms to measure the extent of poor management practices, and whether they can be changed. Another control sample was also monitored. Firms receiving consulting advice experienced large improvements in monitoring systems, quality control, and human resource practices. Average productivity rose by 11 percent through improved quality and efficiency and reduced inventory. And the adoption of decentralized decision making led to better information flows and enabled owners to delegate more day-to-day decisions to middle managers. Many of these practices were immediately profitable, so why had firms bypassed such low hanging fruit? The study finds that informational barriers were at play, and that constraints on firm entry and growth prevent badly managed firms from being driven from the market. Both types of constraints can be alleviated by policies that ensure better business education, a more vibrant local consulting industry, and the removal of barriers to multinational entry, which can potentially help reduce information barriers. Policies can also help reduce corruption and ease financing constraints, making it easier for better- managed firms to expand. 107 62. All this research work is feeding into the 2013 World Development Report and the Knowledge Platform on Jobs, which focuses on job creation, taking a multi-sectoral approach. Prepared by Asli Demirgüç-Kunt, Senior Research Manager, Development Research Group, Development Economics Vice Presidency. 104 de Mel, Suresh, David McKenzie, and Christopher Woodruff. 2008. “Returns to Capital in Microenterprises: Evidence from a Field Experiment.� Quarterly Journal of Economics 123(4): 1329–72. 105 de Mel, Suresh, David McKenzie, and Christopher Woodruff. Forthcoming. “Are Women More Credit Constrained? Experimental Evidence on Gender and Microenterprise Returns.� American Economic Journal: Applied Economics. (Based on Policy Research Working Paper 4746, World Bank, Washington, DC.) 106 McKenzie, David, and Yaye Seynabou Sakho. Forthcoming. “Does It Pay Firms to Register for Taxes? The Impact of Formality on Firm Profitability.� Journal of Development Economics. (Based on Policy Research Working Paper 4449, World Bank, Washington, DC.) 107 Bloom, Nicholas, Benn Eifert, Aprajit Mahajan, David McKenzie, and John Roberts. 2011. “Does Management Matter? Evidence from India.� Policy Research Working Paper 5573, World Bank, Washington, DC. 82 A.4 Human Development and Public Services Research Program 63. Human development and public service delivery are at the core of the World Bank’s strategy to improve people’s lives and support sustainable development. This research program spans education, health, social protection and labor, as well as governance issues affecting service delivery, and the effectiveness of development assistance. Themes 64. This research program continues to cover the full gamut of human development— education, health, labor markets, and social protection. It examines the performance of these sectors in terms of levels and inequalities in access, use, quality, and outcomes. It also examines the impacts of measures to improve performance aimed at households (such as conditional cash transfer and health insurance programs); service providers (including payment systems, legal institutions, and community monitoring mechanisms); policy makers and politicians; and donors (including factors affecting aid and aid’s impact on outcomes). Highlights New “stylized facts� 65. Research using 84 Demographic Health Surveys finds that—in contrast to child mortality— adult mortality in Sub-Saharan Africa has been rising, with the increase heavily concentrated among males, and in countries most affected by HIV/AIDS. 108 Related research finds no consistent pattern in Sub-Saharan African countries of HIV/AIDS prevalence across socioeconomic groups. 109 However, one consistent pattern that does emerge: remarried people have higher-than-average HIV prevalence 110; these individuals form a large portion of the population, and typically outnumber the divorced, separated, and widowed, the groups that have hitherto been the focus of preventive efforts. 66. New research documents high levels of ‘son preference’ in South Asia but also in eastern Europe and central Asia. 111 The same research shows that better-educated women in South Asia do not exhibit lower levels of son preference as manifested in continued childbearing after the birth of girls. Other research gives grounds for hope that the “daughter deficit� in Asia may be 108 de Walque, Damien, and Deon Filmer. 2011. “Trends and Socioeconomic Gradients in Adult Mortality around the Developing World.� Policy Research Working Paper 5716, World Bank, Washington, DC. 109 Beegle, Kathleen G., and Damien de Walque. 2009. “Demographic and Socioeconomic Patterns of HIV/AIDS Prevalence in Africa.� In The Changing HIV/AIDS Landscape: Selected Papers for The World Bank’s Agenda for Action in Africa, 2007-2011, ed. E.L. Lule, R.M. Seifman, and A.C. David, 81–104. Washington, DC: World Bank. 110 De Walque, Damien, and Rachel Kline. 2009. “The Association between Remarriage and HIV Infection: Evidence from National HIV Surveys in Africa.� Policy Research Working Paper 5118, World Bank, Washington, DC. 111 Filmer, Deon, Jed Friedman, and Norbert Schady. 2008. “Development, Modernization, and Son Preference in Fertility Decisions.� Policy Research Working Paper 4716, World Bank, Washington, DC. 83 declining: one paper documents the reduction over time in the sex ratio imbalance in Korea, 112 while another finds an apparent peaking of the imbalance in China and India. 113 67. New research challenges the view that populations, especially poor ones, underuse health services. 114 It documents the high rates of use and access to ambulatory care providers in India, even among the poor; often the doctors in question are not public ones, but rather private, and often untrained. The bigger challenge, the research suggests, is raising the quality of care used, not its quantity. Here too the research comes up with surprises, with doctors in public clinics providing very poor quality care, with the better quality care being found in public hospitals and private clinics. 68. Despite the widespread agreement that they matter more for economic growth than school enrollment, there is a shortage of data for developing countries on cognitive skills. New research highlights the huge variation within India in mathematics achievement; only South Africa records such a large variation. 115 The worst-performing 5 percent of children in Orissa and Rajasthan fare badly by international standards, but the best-performing children fare well by international standards. Household behaviors, human development outcomes, risk and shocks 69. Unsurprisingly, much of the work on households during the last couple of years has concerned economics shocks. Several papers look at their effects on households. One using a large data set of 1.7 million births in 59 developing countries, finds a large, negative association between per capita GDP and infant mortality, with female mortality being more sensitive than male mortality to negative economic shocks. 116 Another explores the psychological impacts of the 1997 Indonesian financial crisis. 117 The study finds substantial increases in several dimensions of psychological distress among men and women across the age distribution, with especially pronounced impacts among vulnerable groups, including those with low education, the rural landless, urban residents, and those in provinces most affected by the crisis. Elevated psychological distress persisted even after the economy returned to pre-crisis levels. In contrast, another study finds that while the labor market effects of the recent crisis in China were large, 112 Chun, Heeran, and Monica Das Gupta. 2009. “Gender Discrimination in Sex Selective Abortions and Its Transition in South Korea.� Womens Studies International Forum 32(2): 89–97. 113 Das Gupta, Monica, Woojin Chung, and Li Shuzhuo. 2009. “Evidence for an Incipient Decline in Numbers of Missing Girls in China and India.� Population and Development Review 35(2): 401–16. 114 Das, Jishu. 2011. “The Quality of Medical Care in Low-Income Countries: From Providers to Markets.� PLOS Medicine 8(4): 1–2. 115 Das, Jishu, and Tristan Zajonc. 2010. “India Shining and Bharat Drowning: Comparing Two Indian States to the Worldwide Distribution in Mathematics Achievement.� Journal of Development Economics 92(2): 175–87. 116 Baird, Sarah, Jed Friedman, and Norbert Schady. 2011. “Aggregate Income Shocks and Infant Mortality in the Developing World.� Review of Economics and Statistics 93(3): 847–56. 117 Friedman, Jed, and Duncan Thomas. 2009. “Psychological Health before, during, and after an Economic Crisis: Results from Indonesia, 1993-2000.� World Bank Economic Review 23(1): 57–76. 84 they were short-lived, with half of the 49 million workers laid off being rehired within six months. 118 70. The economic crisis led to concerns about likely reductions in remittances from migrant workers overseas; research has shed light on the human development effects of migration. One study explores the impact of migration from Guatemala to the United States on height-for-age Z (HAZ) scores and stunting prevalence of children. HAZ scores for children in households with a migrant to the United States are conservatively estimated to be 0.5 standard deviations higher and the prevalence of stunting is approximately 6 percentage points lower. 119 The authors suggest the reasons are likely to be improved food security and reduced morbidity. Two papers take advantage of the lottery scheme that New Zealand operates for would-be Tongan immigrants to get round the biases introduced by nonrandom selection of migrants. One paper finds that migration leads to improvements in mental health, particularly for women and those with poor mental health. 120 Another looks at the effects on children who migrate from Tonga to New Zealand and on those from the same family who are left behind in Tonga: it finds that diets and malnutrition indicators diverge upon migration, with the leavers experiencing improvements and the stayers seeing deteriorations. 121 71. How to cushion the effects of shocks and reduce risk? Several publications examine how public programs can protect children’s nutrition. One provides an overview of the key issues. 122 Another documents the efforts in Indonesia in the 1997–98 crisis; exploiting the differential exposure across communities and children, the paper estimates that the program improved the nutritional status of children 12-24 months of age at the time of the survey, and helped avoid problems of severe malnutrition among young children. 123 Another compares two school- feeding schemes in Burkina Faso, 124 while another analyzes the impacts of a program in Laos; the studies found mixed impacts on nutrition and school enrollment. 125 118 Huang, Jikun, Huayong Zhi, Zhurong Huang, Scott Rozelle, and John Giles. 2011. “The Impact of the Global Financial Crisis on Off-farm Employment and Earnings in Rural China.� World Development 39(5): 797–807. 119 Carletto, Calogero, Katia Covarrubias, and John A. Maluccio. 2011. “Migration and Child Growth in Rural Guatemala.� Food Policy 36(1): 16–27. 120 Stillman, Steven, David McKenzie, and John Gibson. 2009. “Migration and Mental Health: Evidence from a Natural Experiment.� Journal of Health Economics 28(3): 677–87. 121 Gibson, John, David McKenzie, and Steven Stillman. 2011. “What Happens to Diet and Child Health When Migration Splits Households? Evidence from a Migration Lottery Program.� Food Policy 36(1): 7–15. 122 de Pee, Saskia, Henk-Jan Brinkman, Patrick Webb, Steve Godfrey, Ian Darnton-Hill, Harold Alderman, Richard D. Semba, Ellen Piwoz, and Martin W. Bloem. 2010. “How to Ensure Nutrition Security in the Global Economic Crisis to Protect and Enhance Development of Young Children and Our Common Future.� J Nutr 140(1): 138S– 42S. 123 Giles, John, and Elan Satriawan. 2010. “Protecting Child Nutritional Status in the Aftermath of a Financial Crisis: Evidence from Indonesia.� Policy Research Working Paper 5471, World Bank, Washington, DC. 124 Kazianga, Harounan, Damien de Walque, and Harold Alderman. 2009. “Educational and Health Impacts of Two School Feeding Schemes: Evidence from a Randomized Trial in Rural Burkina Faso.� Policy Research Working Paper 4976, World Bank, Washington, DC. 125 Buttenheim, Alison, Harold Alderman, and Jed Friedman. 2011. “Impact Evaluation of School Feeding Programs in Lao PDR.� Policy Research Working Paper 5518, World Bank, Washington, DC. 85 72. Conditional cash transfer (CCT) programs continue to be a much-discussed policy instrument to raise levels of school enrollment and use of preventive care among the chronically poor, as well as to protect vulnerable households at times of crisis. A widely cited Policy Research Report concludes that CCT programs have reduced short-term poverty, and increased enrollment in education and use of health services, but have had only very modest impacts on education and health outcomes. 126 73. Subsequent Bank research has been broadly consistent with these findings. A school stipend program in Punjab, Pakistan, targeted on girls raised enrolments by 9 percent. 127 [20]; subsequent research found that the effects persisted five years into the program. 128 A scholarship program in Cambodia increased attendance by approximately 25 percentage points among the target population (poor households). 129 But research also illustrates the limits to CCT impacts. Eighteen months after the scholarships were awarded in Cambodia, the recipient children did no better on mathematics and vocabulary tests than they would have done in the absence of the program; self-selection by lower-ability students into school as a result of the program appears to explain the lack of the program’s impact on test scores. 130 Research on a CCT program in Colombia also failed to find positive (long-term) impacts on test scores. 131 74. Unsurprisingly, transfers have been found to be subject to diminishing returns: these were evident in the Cambodia program even though the transfers in the program are equivalent on average to only 3 percent of the consumption of the median recipient household. 132 CCT programs can have beneficial effects on outcomes other than those that are the focus of the program: for example, a CCT program aimed at keeping young Malawian women in school led to significant declines in early marriage, teenage pregnancy, and self-reported sexual activity among program beneficiaries after just one year of program implementation. 133 The design of a CCT program can be adjusted to improve outcomes and lower costs: for example, making the 126 Fiszbein, Ariel, Norbert Schady, with Francisco H.G. Ferreira, Margaret Grosh, Niall Keleher, Pedro Olinto, and Emmanuel Skoufias. 2009. Conditional Cash Transfers: Reducing Present and Future Poverty. Policy Research Report. Washington, DC: World Bank. 127 Chaudhury, Nazmul, and Dilip Parajuli. 2010. “Conditional Cash Transfers and Female Schooling: The Impact of the Female School Stipend Programme on Public School Enrolments in Punjab, Pakistan.� Applied Economics 42(28-30): 3565–83. 128 Alam, Andaleeb, Javier E. Baez, and Ximena V. Del Carpio. 2011. “Does Cash for School Influence Young Women’s Behavior in the Longer Term? Evidence from Pakistan.� Policy Research Working Paper 5669, World Bank, Washington, DC. 129 Filmer, Deon, and Norbert Schady. 2011. “Does More Cash in Conditional Cash Transfer Programs Always Lead to Larger Impacts on School Attendance?� Journal of Development Economics 96(1): 150–57. 130 Filmer, Deon, and Norbert Schady. 2009. “School Enrollment, Selection and Test Scores.� Policy Research Working Paper 4998, World Bank, Washington, DC. 131 Baez, Javier E., and Adriana Camacho. 2011. “Assessing the Long-Term Effects of Conditional Cash Transfers on Human Capital: Evidence from Colombia.� Policy Research Working Paper 5681, World Bank, Washington, DC. 132 Filmer, Deon, and Norbert Schady. 2011. “Does More Cash in Conditional Cash Transfer Programs Always Lead to Larger Impacts on School Attendance?� Journal of Development Economics 96(1): 150–57. 133 Baird, Sarah, Ephraim Chirwa, Craig McIntosh, and Berk Özler. 2010. “The Short-Term Impacts of a Schooling Conditional Cash Transfer Program on the Sexual Behavior of Young Women.� Health Economics 19: 55–68. 86 transfer partially conditional on reenrollment or graduation makes sustained increases in enrollment more likely. 134 75. Expanding formal health insurance to groups currently covered through tax-funded public facilities continues to be a widely discussed policy in the developing world. Research from around the Bank points to mixed results. In some countries, such as China, impacts on utilization were found but none were found on out-of-pocket spending. 135 The opposite has been found in other countries, including Georgia 136 and Vietnam. 137 The shift in east Europe and central Asia during the 1990s from tax-financed health “insurance� to the Bismarckian contributory employment-based social health insurance model apparently led to higher rates of spending and utilization of hospital care, but did not lead to lower rates of mortality from conditions amenable to medical care. 138 76. Several papers shed light on the circumstances and vulnerability of the elderly, especially in countries like China where internal migration of younger workers has increased but pension programs are still in their infancy. One shows how the family continues to be an important source of support for the rural elderly, particularly the rural elderly over 70 years of age. 139 It also shows that while the evidence on net financial transfers suggests that elderly with migrant children will receive similar levels of financial transfers as those without migrant children, the predicted variance associated with these transfers implies a higher risk that elderly who have migrant children could fall into poverty. Another paper examines the employment patterns of China’s over-45 population and, for perspective, places them in the context of work and retirement patterns in Indonesia, Korea, the United States, and the United Kingdom. 140 The paper compares workers in the formal urban labor market, under which employees receive generous pensions and face mandatory retirement by age 60, and the informal, predominantly rural, labor market, where workers rely on family support in old age and have much longer working lives. Gender differences in age of exit from work are shown to be much greater in urban China than in 134 Barrera-Osorio, Felipe, Marianne Bertrand, Leigh Linden, and Francisco Perez-Calle. 2011. “Improving the Design of Conditional Transfer Programs: Evidence from a Randomized Education Experiment in Colombia.� American Economic Journal: Applied Economics 3(2): 167–95. 135 Wagstaff, Adam, Magnus Lindelow, Gao Jun, Xu Ling, and Qian Juncheng. 2009. “Extending Health Insurance to the Rural Population: An Impact Evaluation of China’s New Cooperative Medical Scheme.� Journal of Health Economics 28(1): 1–19. 136 Bauhoff, Sebastian, David R. Hotchkiss, and Owen Smith. 2011. “The Impact of Medical Insurance for the Poor in Georgia: A Regression Discontinuity Approach.� Health Economics 20(11): 1362–78. 137 Wagstaff, Adam. 2010. “Estimating Health Insurance Impacts under Unobserved Heterogeneity: The Case of Vietnam’s Health Care Fund for the Poor.� Health Economics 19(2): 189–208. 138 Wagstaff, Adam, and Rodrigo Moreno-Serra. 2009. “Europe and Central Asia’s Great Post-Communist Social Health Insurance Experiment: Aggregate Impacts on Health Sector Outcomes.� Journal of Health Economics 28(2): 322–40. 139 Giles, John, Dewen Wang, and Changbao Zhao. 2010. “Can China’s Rural Elderly Count on Support from Adult Children? Implications of Rural-to-Urban Migration.� Journal of Population Ageing 3(3-4): 183–204. 140 Giles, John, Dewen Wang, and Wei Cai. 2011. The Labor Supply and Retirement Behavior of China’s Older Workers and Elderly in Comparative Perspective.� Policy Research Working Paper 5853, World Bank, Washington, DC. 87 rural areas, and also greater than observed in Korea and Indonesia. Gender differences also emerge in a paper examining how the work, time allocation, and health of nonmigrant women are affected by the out-migration of others in their household. 141 The paper finds that the women left behind are doing more farm work than would have otherwise been the case. There is also evidence that this is a persistent effect, and not just temporary re-allocation. Improving service delivery through stronger accountability 77. The 2004 World Development Report on service delivery highlighted the often poor quality of services in developing countries, and developed a framework for improving performance based on stronger accountability relationships. 142 The report stimulated research in this area, including in the Bank. A new book reviews the recent research on schemes to improve accountability in schools 143; another looks at the role of the judiciary in enforcing rights vis-à-vis service delivery. 144 78. The World Development Report’s “short leg� of accountability emphasized the scope for households to improve provider performance directly, either through the power of their own choices or actions, or through collective action at the community level. Research has explored the scope for citizens to hold frontline providers accountable through grievance and complaint processes. It concludes that, overall, procedures for redressing grievances and complaints regarding basic service delivery are under-developed in many countries, but deserve further analysis, piloting, and support. 145 79. Does collective action offer greater promise? One paper reports the results of a randomized control trail (RCT) of three interventions to encourage beneficiaries’ participation in Indian schools: providing information on existing institutions, training community members in a testing tool for children, and training volunteers to hold remedial reading camps. 146 The interventions had no impact on community involvement, teacher effort, or learning outcomes inside the school. However, in the third intervention, youth volunteered to teach camps, and children who attended substantially improved their reading skills. Another paper reports the 141 Mu, Ren, and Dominique van de Walle. 2009. “Left Behind to Farm? Women’s Labor Re-Allocation in Rural China.� Policy Research Working Paper 5107, World Bank, Washington, DC. 142 World Bank. 2003. World Development Report 2004: Making Services Work for Poor People. Washington, DC: World Bank; Oxford and New York: Oxford University Press. 143 Bruns, Barbara, Deon P. Filmer, and Harry A. Patrinos. 2011. Making Schools Work: New Evidence on Accountability Reforms. Washington, DC: World Bank. 144 Gauri, Varun, and David M. Brinks, eds. 2008. Courting Social Justice: Judicial Enforcement of Social and Economic Rights in the Developing World. New York: Cambridge University Press. 145 Gauri, Varun. 2011. “Redressing Grievances and Complaints Regarding Basic Service Delivery.� Policy Research Working Paper 5699, World Bank, Washington, DC. 146 Banerjee, Abhijit, Rukmini Banerji, Esther Duflo, Rachel Glennerster, and Stuti Khemani. 2010. “Pitfalls of Participatory Programs: Evidence from a Randomized Evaluation in Education in India.� American Economic Journal: Economic Policy 2(1): 1–30. 88 results of a community-based information campaign on school performance from a cluster RCT in 610 villages in three Indian states. 147 Providing information through a structured campaign to communities had a positive impact within four months in all three states. However, the most notable impacts occurred on teacher effort, while impacts on learning were more modest. A third RCT in Indonesia compared different interventions and combinations thereof. 148 Nearly two years on, the study team found that extra grants, measures to reinforce existing school committee structures, and training interventions had led to no perceptible effect. By contrast, test scores had improved in areas that had introduced measures to foster outside ties between the school committee and other parties; in areas that has also introduced elections to the school committee, test scores improved even more. 80. Mass media is often argued to be one mechanism by which citizens can hold politicians accountable for service delivery. A paper using unique data from Benin, finds higher literacy rates among school children in villages exposed to signals from a larger number of community radio stations. 149 However, and in contrast to prior research, the authors find that this media effect does not operate through government accountability: government inputs into village schools and household knowledge of government education policies are no different in villages with greater access to community radio; instead, households with greater access are more likely to make financial investments in the education of their children. Another paper sheds on the question of when politicians exert more effort on behalf of their constituents. 150 The authors examine a unique public spending program that is proliferating across developing countries—the constituency development fund—and find that legislator effort is significantly lower in constituencies where voters are more attached to political parties or reserved for members of socially disadvantaged groups (lower castes). 81. Several papers have looked at the scope for policy makers to improve provider performance through the use of payment methods that link rewards to results. One presents results from a randomized evaluation of a teacher performance pay program implemented across a large representative sample of government-run rural primary schools in the Indian state of Andhra Pradesh. 151 At the end of two years of the program, students in incentive schools performed significantly better than those in control schools by 0.27 and 0.17 standard deviations in math and language tests, respectively. No adverse effects were found, and incentive schools 147 Pandey, Priyanka, Sangeeta Goyal, and Venkatesh Sundararaman. 2009. “Community Participation in Public Schools: Impact of Information Campaigns in Three Indian States.� Education Economics 17(3): 355–75. 148 Pradhan, Menno, Daniel Suryadarma, Amanda Beatty, Maisy Wong, Armida Alishjabana, Arya Gaduh, and Rima Prama Artha. 2011. “Improving Educational Quality through Enhancing Community Participation: Results from a Randomized Field Experiment in Indonesia.� Policy Research Working Paper 5795, World Bank, Washington, DC. 149 Keefer, Philip, and Stuti Khemani. 2011. “Mass Media and Public Services: The Effects of Radio Access On Public Education in Benin.� Policy Research Working Paper 5559, World Bank, Washington, DC. 150 Keefer, Philip, and Stuti Khemani. 2009. “When Do Legislators Pass On ‘Pork’? The Determinants of Legislator Utilization of a Constituency Development Fund in India.� Policy Research Working Paper 4929, World Bank, Washington, DC. 151 Muralidharan, Karthik, and Venkatesh Sundararaman. 2011. “Sundararaman, Teacher Performance Pay: Experimental Evidence from India.� Journal of Political Economy 119(1): 39–77. 89 performed significantly better than other randomly chosen schools that received additional schooling inputs of a similar value. Another paper discusses the scope for improving school performance through the use of contract teachers. 152 Debates over the merit of such teachers have been fierce. Proponents of contract teachers view them as a way to bypass what they see as underperforming regular teachers. Opponents argue that contract teachers unfairly paid less than regular teachers for the same kind of work are subject to arbitrary dismissals and harassment, and do not teach as well as regular teachers. 82. In the health sector, a study reports the results of a randomized evaluation of a program where clinics in Rwanda were paid more for delivering specific preventive interventions and for maintaining high quality-of-care indicators. 153 The scheme had a large and significant positive impact on institutional deliveries and preventive care visits by young children, and improved quality of prenatal care. However, the authors found no effect on the number of prenatal care visits or on immunization rates. The scheme had the greatest effect on those services that had the highest payment rates and needed the lowest provider effort. 83. Another health sector study examines the effects of the shift in Europe and central Asia away from paying hospitals through budgets toward linking payments to activity, either the number and types of services delivered (fee-for-service, or FFS), or the number and type of patients treated (patient-based payments, or PBP). 154 The study found that FFS and PBP both increased national health spending, including private out-of-pocket spending, but had different effects on inpatient admissions (FFS increased them; PBP had no effect), and on average length of stay (FFS had no effect; PBP reduced it). Of the two methods, only PBP appeared to have had any beneficial effect on mortality from causes of death amenable to medical care, but the evidence was weak. Aid and aid effectiveness 84. The economic crisis led to concerns about the supply of development assistance and the demand for loans from the developing world. This research estimates how donor-country banking crises affected aid flows prior to the recent economic crisis. It finds that banking crises in donor countries were associated with a substantial additional fall in aid flows, beyond any income-related effects. Aid flows from crisis-affected countries are estimated to fall by an average of 20 to 25 percent (relative to the counterfactual) and bottom out only about a decade after the banking crisis hits. In addition, the results confirm that donor-country incomes are robustly related to per-capita aid flows, with an elasticity of about three. 155 152 Robinson, Nick, and Varun Gauri. 2011. “Education, Labor Rights, and Incentives: Contract Teacher Cases in the Indian Courts.� Comparative Law and Labour Policy Journal 32 (4): 991–1022. 153 Basinga, Paulin, Paul J. Gertler, Agnes Binagwaho, Agnes L.B. Soucat, Jennifer R. Sturdy, and Christel M. J. Vermeersch. 2010. “Paying Primary Health Care Centers for Performance in Rwanda.� Policy Research Working Paper 5190, World Bank, Washington, DC. 154 Moreno-Serra, Rodrigo, and Adam Wagstaff. 2010. “System-Wide Impacts of Hospital Payment Reforms: Evidence from Central and Eastern Europe and Central Asia. Journal of Health Economics 29(4): 585–602. 155 Dang, Hai-Anh, Stephen Knack, and F. Halsey Rogers. 2009. “International Aid and Financial Crises in Donor Countries.� Policy Research Working Paper 5162, World Bank, Washington, DC. 90 85. Another paper explores countries’ graduation from the International Bank for Reconstruction and Development (IBRD). 156 Graduation is not an automatic consequence of reaching a particular income level, but rather is supposed to be based on a determination of whether the country has reached a level of institutional development and capital-market access that enables it to sustain its own development process. The paper finds that the observed correlates of IBRD graduation are generally consistent with the stated policy. Predicted probabilities generated by the model correspond closely to the actual graduation and de- graduation experiences of most countries (such as Korea and Trinidad and Tobago), and suggest that Hungary and Latvia may have graduated prematurely, a prediction consistent with their subsequent return to borrowing from the Bank in the wake of the global financial crisis. 86. Research also shed light on aid effectiveness. One paper offers new measures of aid quality covering 38 bilateral and multilateral donors, comprising four coherently defined sub- indexes on aid selectivity, alignment, harmonization, and specialization. 157 Compared with earlier indicators used in donor rankings, this indicator set is argued more comprehensive and representative of the range of donor practices addressed in the Paris Declaration, improving the validity, reliability, and robustness of rankings. Another paper looks at the implications of aid fungibility for the evaluation of donor-funded development projects. 158 Using the example of two concurrent World Bank health projects in Vietnam which were targeted on specific provinces, the paper derives impact estimates for different assumptions regarding fungibility and allowing for effects of aid on the productivity of government spending in nonproject provinces. The estimated impacts are highly insensitive to the assumed degree of fungibility, but highly sensitive to the assumed productivity effects. Prepared by Adam Wagstaff, Research Manager, Development Research Group, Development Economics Vice Presidency. A.5 Macroeconomics and Growth Research Program 87. Research on macroeconomics and growth focuses on understanding the diversity of aggregate economic performance across the world, and how it responds to policy and institutional changes under different country circumstances, in order to design policies and reform strategies conducive to sustained high growth, the essential ingredient of lasting poverty reduction. 156 Heckelman, Jac C., Stephen Knack, and F. Halsey Rogers. 2011. “Crossing the Threshold: An Analysis of IBRD Graduation Policy.� Policy Research Working Paper 5531, World Bank, Washington, DC. 157 Knack, Stephen, F. Halsey Rogers, and Nicholas Eubank. 2011. “Aid Quality and Donor Rankings.� World Development 39(11): 1907–17. 158 Wagstaff, Adam. 2011. “Fungibility and the Impact of Development Assistance: Evidence from Vietnam’s Health Sector.� Journal of Development Economics 94(1): 62–73. 91 Themes 88. Research on economic growth investigates how micro- and macroeconomic policy actions and reforms translate into lasting growth, with attention to the role of country-specific initial conditions and policy complementarities. 89. Research on macroeconomic risk management seeks to understand the sources of macroeconomic instability and the mechanisms of its propagation across countries, as well as the consequences for growth and welfare. The research focuses on identifying policy strategies to deal with aggregate risk. Of particular interest are the lessons from the global crisis for managing risk under deepening international integration. 90. Research on governance and political economy examines the consequences of weak governance for government performance, macroeconomic stability and growth, and identifies institutional reforms and incentives that contribute to improving governance. Highlights After the global crisis: new questions on macro-financial policies and propagation mechanisms 91. The global crisis has prompted a reassessment of the macroeconomic and development policies adopted in recent years by many developing economies. The macroeconomic dimensions of financial fragility, ignored by traditional micro-prudential regulation, call for the design of an adequate macro-prudential regime to deter boom-bust cycles of credit and asset prices. Macroeconomic policies need to assist in this task by adopting a more countercyclical stance. In particular, pending the development of suitable macro-prudential tools, monetary policy may face a challenging balance between its primary objective of price stability and the need to safeguard financial stability. 159,160 92. Financial imbalances are typically built up in the upswing, largely because economic bonanzas tend to exacerbate financial market imperfections. Financial systems have limited ability to screen good projects from bad ones, and favorable aggregate shocks encourage entry of high-risk but low-productivity projects. This limits the beneficial effects of such shocks and amplifies their contribution to aggregate financial fragility. 161 The adverse consequences materialize in bad times, in which small macroeconomic shocks can quickly cause big financial turmoil. Evidence from bank runs in Argentina and Uruguay in the 2000s shows that macroeconomic risk affects deposits quite aside from the bank-specific risk characteristics 159 Demirgüç-Kunt, Asli, and Luis Servén. 2010. “Are All the Sacred Cows Dead? Implications of the Financial Crisis for Macro- and Financial Policies.� World Bank Research Observer 25(1): 91–124. 160 de la Torre, Augusto, Alain Ize, and Sergio L. Schmukler. 2011. Financial Development in Latin America and the Caribbean: The Road Ahead. Washington, DC: World Bank. 161 Hoff, Karla. 2010. “Dysfunctional Finance: Positive Shocks and Negative Outcomes.� Journal of Globalization and Development 1(1): Article 4, January. 92 underscored by conventional analyses. Differences in bank exposure to macroeconomic factors can explain differences in deposit withdrawals across banks, especially during turbulent times. 162 93. The crisis has called attention to the mechanisms underlying the amplification and international propagation of financial turbulence, which have been strengthened by global integration. Indeed, close inspection of trade and financial data reveals that the main channels of contagion across countries—international trade linkages, and financial linkages through the cross-border asset holdings of international investors— are stronger today than in the 1990s, thus enlarging the potential for rapid propagation of shocks. 163,164 94. Trade credit represents a potentially important—but often neglected—propagation channel, as a growing number of nonfinancial firms simultaneously receive credit from their suppliers and grant it to their customers. International evidence reveals that increased use of this kind of credit along a supply chain linking different industries significantly increases the correlation between their outputs, which shows that trade credit propagates shocks across industries. 165 95. Banks’ reliance on wholesale funding provided a major propagation mechanism during the global crisis. Wholesale markets dried up, as lenders staged what in effect amounted to a run on banks. An assessment of the impact of the liquidity crunch that followed the demise of Lehman Brothers uncovers a large return differential, both globally and within countries, between the stocks of banks more reliant on nondeposit sources of funds and those less dependent on such funding. 166 This echoes the long-held view that short-term borrowing represents a source of vulnerability to financial shocks, and begs the question of why developing countries borrow short term. The simple answer, based on detailed data on sovereign bond prices and issuances at different maturities, is that it is cheaper than borrowing long term. Countries typically pay a higher risk premium on longer-term bonds, especially in times of crisis. This can be viewed as the outcome of a risk-sharing problem between an emerging economy subject to rollover crises, and risk-averse international investors. 167 162 Levy-Yeyati, Eduardo, María Soledad Martínez Pería, and Sergio L. Schmukler. 2010. “Depositor Behavior under Macroeconomic Risk: Evidence from Bank Runs in Emerging Economies.� Journal of Money, Credit, and Banking 42(4): 585–614. 163 Schmukler, Sergio L., Tatiana Didier, and Paolo Mauro. 2008. “Vanishing Contagion?� Journal of Policy Modeling 30(5): 775–91. 164 Aizenman, Joshua, and Brian Pinto. Forthcoming. “Managing Financial Integration and Capital Mobility: Policy Lessons from the Past Two Decades.� Review of International Economics. (Based on Policy Research Working Paper 5786, World Bank, Washington, DC.) 165 Raddatz, Claudio E. 2010. “Credit Chains and Sectoral Comovement: Does the Use of Trade Credit Amplify Sectoral Shocks?� Review of Economics and Statistics 92(4): 985–1003. 166 Raddatz, Claudio E. 2010. “When the Rivers Run Dry: Liquidity and the Use of Wholesale Funds in the Transmission of the US Subprime Crisis.� Policy Research Working Paper 5203, World Bank, Washington, DC. 167 Broner, Fernando, Guido Lorenzoni, and Sergio L. Schmukler. Forthcoming. “Why Do Emerging Economies Borrow Short Term?� Journal of the European Economic Association. (Based on Policy Research Working Paper 3389, World Bank, Washington, DC.) 93 96. Mutual funds have become increasingly important players in global financial markets, and evidence shows that they engaged in large portfolio reallocations during the global crisis. Their behavior tends to be pro-cyclical, reducing their exposure to countries during bad times and increasing it when conditions improve. Hence capital flows from mutual funds do not seem to play a stabilizing role, but rather expose the countries in their portfolios to foreign shocks. 168 97. Temporary capital controls might help shelter countries from external financial shocks accruing through these channels, but evidence on their effectiveness has remained elusive. Yet the fact that major emerging-market stocks now trade in both domestic and international markets this allows a direct assessment of the effects of capital controls. In their absence, price deviations across markets are rapidly arbitraged away, particularly for liquid stocks. But barriers to cross- border capital flows effectively segment markets: controls on capital outflows induce positive premia between domestic and foreign markets, while controls on inflows have the opposite effect. The size of these premia varies with the intensity of capital controls. 169 98. The effect of the global crisis on different countries entails lessons for macroeconomic risk management. 170 Contrary to popular perception, emerging-market economies suffered growth collapses comparable to, or even larger than, those of advanced economies in 2008–09. But the former recovered faster than the latter and resumed growth at a faster rate post-crisis. Moreover, the policy response of emerging economies represented a sharp departure from the past: this time they implemented countercyclical macro-financial policies and managed to cushion the impact of the crisis. This was facilitated by more prudent fiscal and monetary policies prior to the crisis, as well as by the increased exchange rate flexibility that many countries allowed in response to shocks. 171,172 99. Macroeconomic crises often leave countries saddled with a “debt overhang� that makes them vulnerable to financial shocks and deters growth. In recent years, this has prompted debt relief measures for poor countries. But the economic benefits of debt relief have been difficult to sort out, partly because expected economic performance drives the decision to grant relief. An event study of the response of the stock prices of South African multinationals with subsidiaries in recipient countries to the announcement of debt relief initiatives helps resolve these “reverse causality� problems. The stock prices of these companies exhibit a significant increase above those of other firms, suggesting that financial markets view debt relief as beneficial for firms 168 Raddatz, Claudio E., and Sergio L. Schmukler. Forthcoming. “On the International Transmission of Shocks: Micro-Evidence from Mutual Fund Portfolios.� Journal of International Economics. (Based on NBER Working Paper 17358, Cambridge, Mass.) 169 Levy-Yeyati, Eduardo, Sergio L. Schmukler, and Neeltje Van Horen. 2009. “International Financial Integration through the Law of One Price: The Role of Liquidity and Capital Controls.� Journal of Financial Intermediation 18: 432–63. 170 Ghosh, Swati, Naotaka Sugawara, and Juan Zalduendo. 2011. “Finding a Balance Between Growth and Vulnerability Trade-offs.� Policy Research Working Paper 5592, World Bank, Washington, DC. 171 Didier, Tatiana, Constantino Hevia, and Sergio L. Schmukler. 2011. “How Resilient Were Emerging Economies to the Global Crisis?� Policy Research Working Paper 5637, World Bank, Washington, DC. 172 Schmukler, Sergio L., Tatiana Didier, and Constantino Hevia. 2011. “Emerging Country Responses to the Global Crisis.� In Financial Contagion: The Viral Threat to the Wealth of Nations, ed. Robert W. Kolb. New York: Wiley. 94 operating in recipient countries. This is consistent with an expected improvement in economic conditions, as implied by the “debt overhang� argument for debt relief. 173 100. In many countries, expansionary fiscal policy played a central role in the response to the global crisis. But there is little solid evidence on the magnitude of its effects, because identifying them requires a strategy for finding changes in government spending that are themselves not driven by current macroeconomic shocks. In aid-dependent, low-income countries, one way to do so is by exploiting the long lags between approval and full disbursement of World Bank– financed projects. Nearly all of World Bank–financed government spending in a given year reflects project approval decisions made in previous years, and so does not react to current macroeconomic shocks. Using this strategy, the impact of an additional dollar of government spending on output is estimated to be zero, which suggests the need for considerable skepticism regarding the effectiveness of short-run, spending-based fiscal stimulus packages in poor countries. 174 Microeconomic regulation is a major determinant of macroeconomic performance 101. There is broad agreement that productivity improvements account for the bulk of long- run income growth. Much less is known—notably in developing countries—about how firm- level productivity contributes to aggregate efficiency and growth. A key mechanism seems to be the Schumpeterian “creative destruction� process that transfers resources from low- to high- productivity industries and firms, but its role in emerging economies has remained relatively unexplored. New research examines the regulatory framework faced by firms in developing countries and the implications for firm renewal and macroeconomic performance. 175 102. Excessive or inadequate microeconomic regulation of goods and factor markets limits the reallocation of resources and distorts both the relative profitability of different activities and the incentives to innovation, which is a key driver of growth. 176,177 By limiting the economy’s ability to redeploy resources in the face of shocks, this leads to higher aggregate volatility as well. Moreover, the adverse effects of excessive regulation on growth and stability are particularly large under poor-governance conditions. Thus, streamlining the regulatory framework is likely to bring significant social payoffs especially in developing countries burdened by weak governance. 173 Raddatz, Claudio E. 2011. “Multilateral Debt Relief through the Eyes of Financial Markets.� Review of Economics and Statistics 93: 1262–88. 174 Kraay, Aart C. Forthcoming. “How Large is the Government Spending Multiplier? Evidence from World Bank Lending.� Quarterly Journal of Economics. (Based on Policy Research Working Paper 5500, World Bank, Washington, DC.) 175 Loayza, Norman V., and Luis Servén. 2010. Business Regulation and Economic Performance. Washington, DC: World Bank. 176 Bergoeing, Raphael, Norman V. Loayza, and Facundo Piguillem. 2010. “Why Are Developing Countries So Slow in Adopting New Technologies? The Aggregate and Complementary Impact of Micro Distortions.� Policy Research Working Paper 5393, World Bank, Washington, DC. 177 Brahmbhatt, Milan, and Albert Hu. 2010. “Ideas and Innovation in East Asia.� World Bank Research Observer 25(2): 177–207. 95 103. In turn, evidence from Indonesia shows that deep crises unambiguously weaken the reallocation process. The East Asian crisis attenuated the relationship between productivity and survival, causing the exit of relatively productive firms. On the bright side, firms that entered during the crisis were relatively more productive, which helped mitigate the fall in aggregate productivity. 178 The growth payoff of reforms depends on complementary measures and initial conditions 104. A considerable literature has examined the merits of reform packages vis-à-vis specific reforms. International evidence for the case of trade liberalization shows that the former option is likely to deliver a bigger growth impact. The reason is that the ability of reforming countries to take advantage of increased international competition and access to foreign markets depends on complementary reforms and structural factors that shape the availability of productive inputs—such as human capital and infrastructure services—and the efficiency with which they can be reallocated to their most productive uses following trade reform, as determined by labor market flexibility, financial market development, overall governance quality, and ease of firm entry and exit. The growth payoff from trade reform is bigger when the removal of obstacles to trade is accompanied by complementary domestic reforms in these areas. 179 Related research shows that, given the set of accompanying measures, trade reform has a bigger growth impact when prior to reform trade is severely distorted—namely, the marginal benefit declines with the extent of reform. 180 The effect of disasters on growth depends on the type of disaster and the stage of economic development 105. Despite the tremendous human suffering caused by natural disasters, previous research has not reached clear conclusions regarding their effects on economic growth. Closer analysis distinguishing among disaster types and economic sectors yields three major insights. First, the effect of disasters on economic growth is not always negative—droughts, for instance, tend to reduce agricultural growth, while floods have a positive effect on overall growth. Second, although moderate disasters can have a positive growth effect in some sectors, severe disasters do not. Third, growth in developing countries is more sensitive to natural disasters than in 178 Hallward-Driemeier, Mary, and Bob Rijkers. 2011. “Do Crises Catalyze Creative Destruction? Firm-level Evidence from Indonesia.� Policy Research Working Paper 5869, World Bank, Washington, DC. 179 Chang, Roberto, Linda Kaltani, and Norman V. Loayza. 2009. “Openness Can Be Good for Growth: The Role of Policy Complementarities.� Journal of Development Economics 90(1): 33–49. 180 Loayza, Norman V., and Linda Kaltani. 2008. “Initial Conditions and the Outcome of Economic Reform.� Economics Letters 101(3): 230–33. 96 developed ones, with more sectors affected and the effects larger and economically meaningful. 181 106. Policy can have a significant impact on the mortality caused by disasters. In the case of earthquakes, governments can enforce quakeproof construction regulation, but many do not. In poor countries in which earthquakes are rare, households and governments are reluctant to divert resources from pressing immediate needs to costly measures to prevent earthquake mortality. But even in places where earthquakes are more common, the government response also depends on incomes and political incentives of governments to provide public goods. Governments with fewer political incentives to provide public goods respond less to an increase in earthquake propensity. This means that post-earthquake relief must remain a high priority for donors. 182 Collective action by informed citizens and good governance may be mutually reinforcing 107. There is broad consensus that good governance contributes to improved growth and development outcomes, and collection and update of suitable measures of governance remains a research priority. 183,184 However, less is known about how to achieve good governance. Citizens’ ability to enforce policy makers’ accountability is viewed by many as a key to improve governance. Political parties can play an important role in this regard, by allowing citizens to take collective action to sanction political actors that renege on their commitments. Across countries, there is a robust positive relation between development outcomes and the degree to which political parties facilitate collective action. This underscores the need for greater attention to the organizational characteristics of countries that allow citizens to hold leaders accountable. 185 108. A similar mechanism is partly responsible for the fact that democratizing countries typically display worse fiscal outcomes—lower public investment and greater corruption—than established democracies. The difference is due to the low credibility of electoral promises in new democracies, which leads to higher targeted transfers and corruption and lower public good provision. 186 In turn, research using data from India suggests that citizen information can 181 Loayza, Norman V., Eduardo Olaberría, Jamele Rigolini, and Luc Christiaensen. Forthcoming. “Natural Disasters and Growth: Going beyond the Averages.� Journal of Applied Econometrics. (Based on Policy Research Working Paper 4980, World Bank, Washington, DC.) 182 Keefer, Philip E., Eric Neumayer, and Thomas Pluemper. 2011. “Earthquake Propensity and the Politics of Mortality Prevention.� World Development 39 (9): 1530–41. 183 Kraay, Aart, and Dani Kaufmann. 2008. “Governance Indicators: Where Are We, and Where Should We Be Going?� World Bank Research Observer 23(Spring): 1–30 184 A major output from this work is the widely used Worldwide Governance Indicators dataset covering more than 200 countries since 1996, available at http://www.govindicators.org. 185 Keefer, Philip E. 2011. “Collective Action, Political Parties and Pro-Development Public Policy.� Asian Development Review 28: 94–118. 186 Keefer, Philip E., and Razvan Vlaicu. 2008. “Democracy, Credibility and Clientelism.� Journal of Law, Economics and Organization 24(2): 371–406. 97 facilitate collective action and thereby help improve governance. India’s constituency development funds allow individual legislators to finance local public infrastructure in their electoral districts. The data show that politicians spend less of their fund entitlement when media coverage of the program is low. 187 109. Unaccountable institutions may themselves undermine collective action by citizens. Research using data from the Gallup World Poll documents a quantitatively large and statistically significant negative correlation between corruption and confidence in public institutions. The correlation is robust to the inclusion of a large set of controls for country and respondent-level characteristics, so it reflects at least in part a causal effect of corruption on confidence. Individuals with low confidence in institutions exhibit low levels of political participation, show increased tolerance for violent means to achieve political ends, and have a greater desire to “vote with their feet� through emigration. 188 Overall, this suggests that corruption can inhibit development by eroding confidence in public institutions. 110. Individuals’ inability to cooperate with one another, and to enforce social norms that restrain opportunistic behavior, can deter development by weakening government accountability, property rights, and contract enforcement. Results from experiments conducted in India to study coordination and to measure social preferences reveal that a large majority of subjects exhibit “spiteful preferences�: they punish cooperative behavior, even though such punishment increases inequality, decreases the payoffs of all participants, and weakens the deterrent effect of punishment of norm violations. Further, the willingness to reduce another’s material payoff— either for the sake of achieving more equality or for the sake of being ahead—is stronger among individuals belonging to high castes than among those belonging to low castes. 189 Related experiments show that individuals at the bottom of the caste hierarchy also exhibit a much lower willingness than the rest to punish norm violations that hurt members of their own caste, 190 which may hamper their prospects for overcoming poverty and likely reflects the effect on their social preferences of a long-standing legacy of inequality and social exclusion. 191 Prepared by Luis Servén, Senior Advisor, Development Research Group, Development Economics Vice Presidency. 187 Keefer, Philip E., and Stuti Khemani. 2009. “When do Legislators Pass on Pork? The Role of Political Parties in Determining Legislator Effort.� American Political Science Review 103(1): 99–112. 188 Clausen, Bianca, Aart Kraay, and Zsolt Nyiri. 2011. “Corruption and Confidence in Public Institutions: Evidence from a Global Survey.� World Bank Economic Review 25: 212–49. 189 Fehr, Ernst, Karla Hoff, and Mayuresh Kshetramade. 2008. “Spite and Development.� American Economic Review, Papers & Proceedings 98(2): 494–99. 190 Fehr, Ernst, Karla Hoff, and Mayuresh Kshetramade. 2011. “Caste and Punishment—The Legacy of Caste Culture in Norm Enforcement.� Economic Journal 121(566): F449–75. 191 Fehr, Ernst, and Karla Hoff. 2011. “Tastes, Castes, and Culture: The Influence of Society on Preferences.� Economic Journal 121(556): F396–412. 98 A.6 Poverty and Inequality Research Program 111. When the World Bank speaks of its overarching goal of eliminating poverty, it is understood that the term is being used in the broadest sense; referring not only to raising income levels above some critical threshold, but also to the eradicating hunger, illiteracy, disease, and social exclusion. When this concept is given concrete analytical shape, the initial focus is often on poverty defined in terms of household and individual consumption levels below an appropriate poverty line for the country concerned or a global line, such as the Bank’s international line of $1.25 a day in 2005 prices. Poverty research aims to both deepen and broaden the analysis of poverty with a focus on collecting better data, improving methods to analyze living standards, and understanding better the processes that determine the extent of poverty and the distribution of well-being. Themes 112. Research on poverty and inequality supports the World Bank’s goal of alleviating poverty. Building on insights from the Bank’s flagship World Development Reports on poverty and inequality published during the past two decades, the central research themes have focused on two areas: better data for describing poverty and inequality and research on understanding and breaking poverty and inequality traps. 113. Researchers are active on multiple fronts, producing new household-level survey data, monitoring poverty and inequality using household-level data (including the global poverty monitoring task, which produces the Bank’s official “$1.25 a day� poverty counts), developing improved tools for the analysis of poverty and inequality, and studying in detail some of the processes that govern whether and how individuals are able to escape poverty. As the ongoing global economic crisis continues to evolve, efforts are also focused on gauging its impact on distributional outcomes in the developing world. 114. The Living Standards Measurement Study (LSMS), now in its 30th year, serves as a model for the Bank’s orientation toward open data and open knowledge. The program undertakes research to improve the quality of household-level measurement important for poverty and policy analysis. Through new data collection, methodological experiments, field validations, as well as reviews of existing knowledge the program aims to provide sound advice for improving LSMS and other surveys in developing countries. Highlights Tracking global poverty and inequality 115. DEC’s Poverty and Inequality research program (DECPI) continues to play a central role in the World Bank’s global poverty counts. There has been a dramatic expansion of data underpinning these poverty calculations, and periodic revisions to international purchasing 99 power parity adjustments are also incorporated. A recent study summarizing this effort provides robust evidence of continually declining poverty incidence and depth since the early 1980s, though progress has been uneven across regions and over time. 192 This work is based on a new international poverty line of $1.25 per day at 2005 international purchasing power parity. 193 The importance of such purchasing power parity exchange rates for global poverty estimation is also addressed by researchers across the Bank. 194 116. New measures of global poverty have also been developed, including the “weakly relative measures� which allow the international poverty line to rise with mean consumption when this rises above a critical value, estimated to be about $2 a day. 195 This allows for costs of social inclusion and relative deprivation. The incidence of relative poverty has also been declining in the developing world as a whole, but not sufficiently to bring down the total number of poor. 117. The analysis of global inequality has similarly been receiving attention in DECPI, and as with the poverty estimates, these are based on household survey data corrected for purchasing power variation. 196 The study of global inequality cannot only be updated as new data become available but can also extend back in time for comparative purposes. 197 The focus on inequality at the global level reveals that differences between countries accounts for the lion’s share of overall inequality. The position of an individual in the global income distribution is largely determined by his country of birth. 198 118. The research department has also been studying the economic costs of inequality. A recent study confirmed that high initial inequality impedes subsequent growth and poverty reduction, but that not all inequalities matter; the most important aspect of the initial distribution of income is the extent of absolute poverty. 199 When this is high, it is both harder for an economy 192 Ravallion, Martin, and Shaohua Chen. 2011. “Developing World is Poorer than We Thought, but No Less Successful in the Fight against Poverty.� Quarterly Journal of Economics 125 (4): 1577–1625. 193 Ravallion, Martin, Shaohua Chen, and Prem Sangraula. 2009. “Dollar a Day Revisited.� World Bank Economic Review 23(2): 163–84. 194 Deaton, Angus, and Olivier Dupriez. 2011. “Purchasing Power Parity Exchange Rates for the Global Poor.� American Economic Journal: Applied Economics 3(2): 137–66. 195 Ravallion, Martin and Shaohua Chen. “Weakly Relative Poverty,� Review of Economics and Statistics 93(4): 1251-1261. 196 Milanovic, Branko. Forthcoming. “Global Inequality Recalculated and Updated: The Effect of New PPP Estimates on Global Inequality and 2005 Estimates.� Journal of Economic Inequality. (Based on Policy Research Working Paper 5061, World Bank, Washington, DC.) 197 Lindert, Peter H., Branko Milanovic, and Jeffrey G. Williamsen. 2011. “Pre-Industrial Inequality.� The Economic Journal Volume 121(551, March): 255–72. Milanovic, Branko. 2011. “Global Inequality and the Global Inequality Extraction Ratio: The Story of the Past Two Centuries.� Explorations in Economic History 48 (November): 494-506. 198 Milanovic, Branko. 2011. The Haves and the Have-Nots: A Brief and Idiosyncratic History of Global Inequality. New York: Basic Books. 199 Ravallion, Martin. 2012. “Why Don’t we See Poverty Convergence?� American Economic Review, in press. 100 to grow, and harder to assure that the poor participate fully in that growth. Thus poverty can self- perpetuate despite good economic policies and other favorable initial conditions in human development. A plausible explanation for this finding is that poor people tend to be locked out of access to credit to finance new investments. Drilling down on poverty and the welfare of children 119. A key purpose of poverty analysis is to provide an as accurate a description of the well- being of particular population groups as possible. Such poverty “profiles� inform the design of policy and help to prioritize public actions. One population group of great interest and importance in this connection is children. A large body of analysis has been undertaken by DECPI researchers examining the well-being of children from a variety of perspectives. 120. One study developed a new methodology for measuring the contribution of macroeconomic outcomes—growth and distributional change—to aggregate indicators for child outcomes. 200 Applications for Morocco and Vietnam reveal that growth and distributional change have played only a modest role in the changes in school enrollments over time observed in both countries—pointing instead to the importance of structural changes, consistent with changing attitudes, expectations and supportive public policies to promoting greater schooling. 121. A study of orphanhood in 21 Sub-Saharan Africa countries, prompted by the rise in adult mortality due to HIV/AIDS, notes that grandparents are taking on increased childcare responsibility. 201 Another study, concentrating on northwestern Tanzania, documents that loss of a mother has severe consequences for a child’s health, education, and consumption level. 202 Long-term data on welfare outcomes collected in the Kagera region of Tanzania between 1991 and 2004 reveals that child labor has long-term impacts on education, employment, and marital outcomes but affects boys and girls differently. 203 The impact of the global economic crisis on child survival in Africa provides quantitative estimates of approximately 30,000 to 50,000 excess infant deaths as a result of the crisis, and points to the need to protect incomes of the poor and maintain critical health services. 204 122. A further study looks more generally at the various transmission mechanisms linking economic shocks to child human capital schooling and health. The paper provides evidence that 200 Lambert, Sylvie, Martin Ravallion and Dominique van de Walle. 2010. “A Micro-Decomposition Analysis of Aggregate Human Development Outcomes,�Oxford Bulletin of Economics and Statistics 72(2): 119-145. 201 Beegle, Kathleen, Deon Filmer, Andrew Stokes, and Lucia Tiererova. 2010. “Orphanhood and the Living Arrangements of Children in Sub-Saharan Africa.� World Development 38(12): 1727–46. 202 Beegle, Kathleen, Joachim de Weerdt, and Stefan Dercon. 2010. “Orphanhood and Human Capital Destruction: Is there Persistence into Adulthood?� Demography 47(1): 163–80. 203 Beegle, Kathleen, Rajeev Dehejia, and Roberta Gatti. 2009. “Why Should We Care about Child Labor? The Education, Labor Market, and Health Consequences of Child Labor.� Journal of Human Resources 44(4): 871–89. 204 Baird, Sarah, Jed Friedman, and Norbert Schady. 2011. “Aggregate Income Shocks and Infant Mortality in the Developing World.� Review of Economics and Statistics 93(2): 847–56. 101 impacts can vary markedly between rich and poor countries. 205 Son preference is a widely observed phenomenon in developing countries and can play an important role in shaping fertility decisions. A study based on data from 65 countries shows that son preference is apparent in many regions of the developing world and does not necessarily fall with income growth and modernization. 206 Looking more specifically at the case of India another study indicates that children born during the monsoon months have worse health outcomes than children born during the fall and winter months, reflecting the poorer environmental conditions that such children are generally exposed to. 207 A further study documents that infant and gender differences in mortality vary across Indian population groups defined in terms of religion and caste, in ways that do not necessarily mirror socioeconomic status. 208 Measuring inequality of opportunity 123. The 2006 World Development Report, entitled Equity and Development, placed inequality of opportunity at center stage in development debates. The report identified inequality of opportunity as occurring when there are clear differences in welfare outcomes among population groups defined in terms of characteristics that are “morally irrelevant�—pre- determined characteristics, such as gender, race, place of birth or family background that a given society agrees should have no bearing on current welfare outcomes. Inequality of opportunity can thus be distinguished from the more traditionally measured inequality of outcomes (such as income and consumption) although it is clear that the two are closely related. Measuring inequality has been taken up in various initiatives throughout the World Bank. Notably, in the Bank’s Latin America region there has been research on inequality of opportunity that focuses on inequalities in access to public services. 209 One approach explored by DECPI researchers has been to apply standard techniques of inequality decomposition by population subgroups to survey data. 210 This effort has extended not only to consumption and income inequality 205 Ferreira, Francisco H. G., and Norbert Schady. 2009. “Aggregate Economic Shocks, Child Schooling and Child Health.� World Bank Research Observer 24(2): 147–81. 206 Filmer, Deon, Jed Friedman, and Norbert Schady. 2009. “Development, Modernization, and Son Preference in Fertility Decisions.� World Bank Economic Review 23(3): 371–98. 207 Lokshin, Michael, and Sergiy Radyakin. 2009. “Month of Birth and Children’s Health in India.� Policy Research Working Paper 4813, World Bank, Washington, DC. 208 Borooah, Vani K., Quy-Toan Do, Sriya Iyer, and Shareen Joshi. 2009. “Missing Women and India’s Religious Demography.� Policy Research Working Paper 5096, World Bank, Washington, DC. 209 World Bank. 2010. Do Our Children Have a Chance? The 2010 Human Opportunity Report for Latin America and the Caribbean. World Bank: Washington, DC. Paes de Barros, Ricardo, Francisco H. G. Ferreira, José R. Molinas Vega, and Jaime Saavedra Chanduvi. 2009. Measuring Inequality of Opportunities in Latin America and the Caribbean. World Bank: Washington, DC. 210 Ferreira, Francisco H. G., and Jérémie Gignoux. 2011. “The Measurement of Inequality of Opportunity: Theory and an Application to Latin America.� Review of Income and Wealth 57(4): 622–57. Aran, Meltem, Francisco H. G. Ferreira, and Jérémie Gignoux. 2011. “Measuring Inequality of Opportunity with Imperfect Data: The Case of Turkey.� Journal of Economic Inequality 9(4): 651–80. 102 decompositions, but has also looked at other welfare outcomes, such as education and labor market earnings. 211 A further effort has been focused on alternative decomposition techniques. 212 Improving and expanding data 124. When the World Bank first attempted to estimate the extent of global poverty in its World Development Report on Poverty in 1990, it drew on 22 household surveys from 22 countries. Today, the Bank’s estimates are based on 700 surveys from 115 countries. Despite this tremendous progress in collecting data many challenges remain. Regional coverage of data remains patchy in Sub-Saharan Africa, a region of concern. Data are often of uncertain quality, and indeed there is much still to be learnt about how to collect data accurately and efficiently. For example, the Integrated Surveys on Agriculture (ISA) project established by the Bill and Melinda Gates Foundation and implemented by the LSMS program in DECPI, is collaborating with national statistics offices in seven countries to design and implement systems of multi-topic nationally representative panel household surveys, with a strong focus on agriculture. 213 125. Alongside generating new data via LSMS-ISA and other data collection activities the DECPI team is researching methods for improving the quality and relevance of information collected. An experimental study of alternative methods of household consumption measurement tests eight alternative methods to measure household consumption on a sample of 4,000 households and points to lessons for consumption measurement in low-income countries. 214 Another study examines the reliability of recall in agricultural data and shows that despite conceptual grounds for concern there is little evidence of recall bias impacting data quality. 215 Yet another study revisits the role of land measurement error in the inverse farm-size productivity relationship, finding that an improved measure of land size using global positioning system information strengthens the evidence in support of the existence of such a relationship. 216 211 Ferreira, Francisco H. G., and Jérémie Gignoux. 2011. “The Measurement of Educational Inequality: Achievement and Opportunity.� Policy Research Working Paper 5873, World Bank, Washington, DC. Bourguignon, François, Francisco H. G. Ferreira, and Phillipe G. Leite. 2008. “Beyond Oaxaca-Blinder: Accounting for Differences in Household Income Distributions.� Journal of Economic Inequality 6(2): 117–48. 212 Elbers, Chris, Peter Lanjouw, Johan Mistiaen, and Berk Özler. 2008. “Re-Interpreting Sub-Group Inequality Decompositions.� The Journal of Economic Inequality 6(3, September): 1569–1721. Lanjouw, Peter, and Vijayendra Rao. 2011. “Revisiting Between-Group Inequality Measurement: An Application to the Dynamics of Caste Inequality in Two Indian Villages.� World Development 39(2): 174–87. 213 Website for Living Standards Measurement Study: Integrated Surveys on Agriculture: www.worldbank.org/lsms-isa. 214 Beegle, Kathleen, Joachim de Weerdt, Jed Friedman, and John Gibson. Forthcoming. “Methods of Household Consumption Measurement through Surveys: Experimental Results from Tanzania.� Journal of Development Economics. (Based on Policy Research Working Paper 5501, World Bank, Washington, DC.) 215 Beegle, Kathleen, Calogero Carletto, and Kristen Himelein. Forthcoming. “Reliability of Recall in Agricultural Data.� Journal of Development Economics. (Based on Policy Research Working Paper 5671, World Bank, Washington, DC.) 216 Carletto, Calogero, Sara Savastano, and Alberto Zezza. 2011. “Fact or Artefact: The Impact of Measurement Errors on the Farm Size Productivity Relationship.� Policy Research Working Paper 5908, World Bank, Washington, DC. 103 126. The team has also been studying the potential for making better use of subjective data in measuring welfare and poverty. Carefully constricted survey questions can help in calibrating multidimensional welfare measures and setting credible poverty lines in specific social contexts. However, users need to be aware of a number of measurement issues, including the potential for “frame-of-reference� bias, whereby different people assume different scales when answering subjective questions on their welfare. 217 This can bias the responses, though one recent study by the team showed that the bias was small, giving encouragement to potential future applications. 218 Getting more out of existing data 127. Notwithstanding the expansion of data many issues and questions that are not easily addressed with existing data remain. Indeed, in many settings the proliferation of datasets is accompanied by new questions associated with data comparability. DECPI research endeavors to contribute to this research agenda by exploring methods to combine existing data so that new evidence can shed light on existing questions. As these methods generally involve estimation, they raise statistical challenges for inference which require close attention. One such example is ongoing research that combines household survey data with population census data with a view to estimating poverty at the small-area level. Two recent studies support the validity of the Bank’s “poverty mapping� method in settings where local poverty rates can be directly calculated and compared to small-area estimates. 219 Another study extends this methodology to the small-area estimation of child malnutrition by combining survey-based data on anthropometric outcomes with population census data. 220 A study in Malawi extends the methodology by combining Demographic and Health Survey and sentinel survey data to produce estimates of HIV prevalence at the local level. 221 Survey to survey imputation methods can also be applied in settings where consumption data comparability is in question or where tracking poverty is hampered by problematic price indices to deflate for cost of living changes. 222 The basic approach has also been applied to develop synthetic panel data for the purpose of tracking 217 Lokshin, Michael and Martin Ravallion. 2010. “Who Cares about Relative Deprivation?� Journal of Economic Behavior and Organization, 73(2): 171-185. 218 Beegle, Kathleen, Kristen Himelein and Martin Ravallion. 2012. “Frame-of-Reference Bias in Subjective Welfare Regressions.� Journal of Economic Behavior and Organization, 81: 556-570. 219 Demombynes, Gabriel, Chris Elbers, Jean O. Lanjouw, and Peter Lanjouw. 2008. “How Good a Map? Putting Small Area Estimation to the Test.� Revista Internazionale di Scienze Sociali 4: 465–93. Elbers, Chris, Peter Lanjouw, and Phillipe G. Leite. 2010. “Brazil within Brazil: Testing the Poverty Map Methodology in Minas Gerais.� Policy Research Working Paper 4513, World Bank, Washington, DC. 220 Fujii, Tomoki. 2010. “Micro-Level Estimation of Child Undernutrition Indicators in Cambodia.� World Bank Economic Review 24(3) 520–53. 221 Ivaschenko, Oleksiy, and Peter Lanjouw. 2010. “A New Approach to Producing Geographic Profiles of HIV Prevalence: An Application to Malawi.� World Medical and Health Policy 2(1): 235–66. 222 Christiaensen, Luc, Peter Lanjouw, Jill Luoto, and David Stifel. Forthcoming. “Small Area Estimation-Based Prediction Methods to Track Poverty: Validation and Applications.� Journal of Economic Inequality. (Based on Policy Research Working Paper 5683, World Bank, Washington, DC.) 104 changes in poverty in settings where only cross section survey data have been collected. 223 Analysis of the resultant synthetic panels can complement that from more conventional pseudo- panels, as produced for example in a recent Bank study for Latin America. 224 Aside from combining data sources, research has also looked at the scope for analyzing economic welfare, such as poverty and inequality, on the basis of asset indices in the large, and growing, number of household surveys that do not collect household income or consumption data. 225 The impact of policies 128. An important theme in DECPI research has been the detailed and careful investigation of the impacts of policy on welfare outcomes—encompassing impacts on income poverty but also health, nutrition, education and other important welfare dimensions. Key to such evaluations is the isolation of policy interventions from other possible causal influences on welfare outcomes. An important research effort has focussed on the impact of a conditional cash transfer (CCT) program in the case of Malawi. This research project evaluates a unique experiment targeted at adolescent girls that featured both a conditional and unconditional treatment arm. The study finds that although there was a modest improvement in both arms relative to the control group, the increase in the unconditional cash transfer was only about half as large as the conditional cash experiment. On the other hand, the study finds that teenage pregnancy and marriage rates were substantially higher in the conditional cash transfer experiment. 226 Further work confirms that the reduction in teenage pregnancies and marriage rates in the unconditional cash transfer arm translated into reductions in risky sexual behaviour and the prevalence of HIV. 227 Another project examined the long-term impact of exposure to a large-scale nutrition program on the nutritional outcomes of children in Madagascar. The study shows that the program protected long-term nutritional outcomes against an underlying negative trend in the absence of the program. 228 A further study evaluates the short-run potential of promoting self-employment activities in Argentina as a means to phase out large safety net programs. The study finds that 223 Lanjouw, Peter, Jill Luoto, and David McKenzie. 2011. “Using Repeated Cross-Sections to Explore Movements In and Out of Poverty.� Policy Research Working Paper 5550, World Bank, Washington, DC. Cruces, Guillermo, Peter Lanjouw, Leonardo Lucchetti, Elizaveta Perova, Renos Vakis, and Mariana Viollaz. 2011. “Intra-Generational Mobility and Repeated Cross-Sections: A Three Country Validation Exercise.� Policy Research Working Paper 5916, World Bank, Washington, DC. 224 Cuesta, José A., and Hugo Nopo, and Georgina Pizzolitto. 2011. “Using Pseudo-Panels to Measure Income Mobility in Latin America.� Review of Income and Wealth 57(2): 224–46. 225 Filmer, Deon, and Kinnon Scott. 2011. “Assessing Asset Indices.� Demography (December). DOI 10.1007/s13524-011-0077-5. 226 Baird, Sarah, Craig McIntosh, and Berk Özler. 2011. “Cash or Condition: Evidence from a Cash Transfer Experiment.� Quarterly Journal of Economics 126(4): 1709–53. 227 Baird, Sarah, Craig McIntosh, and Berk Özler. Forthcoming. “Impact of a Cash Transfer Program for Schooling on Prevalence of HIV and HSV-2 in Malawi: A Cluster Randomized Trial.� The Lancet. Baird, Sarah, Craig McIntosh, and Berk Özler. 2010. “The Short-Term Impacts of a Schooling Conditional Cash Transfer Program on the Sexual Behavior of Young Women.� Health Economics 19(S1): 55–68. 228 Galasso, Emanuela, Nithin Umapathi, and Jeffrey Yau. Forthcoming. “Nutritional Gains from Extended Exposure to a Large Scale Nutrition Programme.� Journal of African Economies (20)5. 105 relatively few households were attracted by this program, but that those that did were more likely to be female-headed households and the well-educated. 229 129. Along with other teams in the research department, an effort is being made to expand policy evaluations to include broader sectoral and economy-wide policies with bearing on poverty and inequality. An example was a study of the poverty impacts of massive agrarian reforms in Vietnam in the 1980s and 1990s, which first entailed privatizing the country’s agricultural land (previously farmed by collectives) and then establishing markets in land-use rights. 230 Using innovative economic modelling approaches drawing on both economic theory and a wide range of data sources, the study revealed both gainers and losers from these reforms, but that on balance the reforms were poverty and inequality reducing, and this was true despite rising landlessness in some rural areas. Insights from qualitative and historical analysis 130. Development policy research needs to look well beyond quantitative analysis of household survey data to the role of institutions in development. The consensus among scholars and policy makers that “institutions matter� for development also means that “history matters�, since institutions clearly form and evolve over time. DECPI researchers have prompted a dialogue between history and development policy by clarifying the role of historical scholarship and offering general principles and implications that historians bring to development policy debates. 231 The focus on community as a key institution of development is at the heart of a penetrating evaluation of community driven development (CDD) projects as a means of achieving poverty reduction in the developing world. This Policy Research Report evaluates whether policies of inducing CDD and the decentralization of resources and authority to local authorities translate into improvements in welfare outcomes. The study suggests civic society failures may hold back the positive impacts from induced participation, and highlights the importance of internal learning and monitoring as a means of improving the performance of community driven projects. 232 Many of the themes raised in the Policy Research Report are analyzed further, and in even greater detail, in a study of participatory projects and local conflict dynamics in Indonesia. 233 Prepared by Peter Lanjouw, Research Manager, Development Research Group, Development Economics Vice Presidency. 229 Galasso, Emanuela, and Rita Almeida. 2010. “Jumpstarting Self-Employment? Evidence among Welfare Participants in Argentina.� World Development 38(5): 742–55. 230 Ravallion, Martin and Dominique van de Walle. 2008. Land in Transition: Reform and Poverty in Rural Vietnam, Palgrave Macmillan. 231 Bayley, C. A., Vijayendra Rao, Simon Szreter, and Michael Woolcock, eds. 2011. History, Historians and Development Policy: A Necessary Dialogue. Manchester, UK: University of Manchester Press. 232 Mansuri, Ghazala, and Vijayendra Rao. Forthcoming. Localizing Development: Does Participation Work? Policy Research Report. Washington, DC: World Bank. 233 Barron, Patrick, Rachael Diprose, and Michael Woolcock. 2011. Contesting Development: Participatory Projects and Local Conflict Dynamics in Indonesia. New Haven: Yale University Press. 106 A.7 Trade and International Integration Research Program 131. Our understanding of country-level trade in goods, and of conventional trade policy— tariffs and quotas—has improved. But other dimensions of international integration, including firm-level transactions and trade in services, and of policies affecting international integration, such as trade facilitation and export promotion, remain less well understood. New research is beginning to remedy the gaps in our knowledge and facilitate informed policy-making. Theme 132. Research focuses on how trade in goods and services, foreign investment and migration, and the policies influencing these flows, affect economic development. Both national policies and international agreements are studied. The specific policy research questions are in many cases defined by demand from Bank operations and client countries; addressed by collecting new data which are rigorously analyzed and widely disseminated; studied in collaboration with researchers from developing countries and other development institutions; and the results implemented in cooperation with staff from Bank operations and client countries. Highlights New ways of measuring and analyzing trade protection are helping us understand its extent and causes 133. Researchers developed three measures of trade distortions, each well-grounded in theory and each capturing a key aspect of trade policy: the Trade Restrictiveness Index (TRI) summarizes the effects of a country’s trade policy on its own welfare; the Overall Trade Restrictiveness index (OTRI) summarizes effects on its import volume, and hence on its trading partners; and the Market Access OTRI (MA-OTRI) summarizes the effects of partner countries’ distortions on its exports. 234 Empirical estimates revealed that poor countries have more restrictive policies, but they also face higher trade barriers on their exports. Including non-tariff barriers (along with tariffs) increases measured trade restrictiveness by an additional 87 percent. 235 A series of studies in Africa are demonstrating how non-tariff barriers also fragment regional markets. 236 Among non-tariff barriers, stringent product standards are becoming a serious hindrance to trade in some sectors. 237 234 Kee, Hiau Looi, Alessandro Nicita, and Marcelo Olarreaga. 2009. “Estimating Trade Restrictiveness Indices.� The Economic Journal 119(153): 172–99. 235 Kee, Hiau Looi, Alessandro Nicita, and Marcelo Olarreaga. 2008. “Import Demand Elasticities and Trade Distortions.� Review of Economics and Statistics 90(4): 666–82. 236 Gillson, Ian. 2010. “Deepening Regional Integration to Eliminate the Fragmented Goods Market in Southern Africa.� Africa Trade Policy Note 9, World Bank, Washington, DC. Kirk, Robert. 2010. “Addressing Trade Restrictive Non-Tariff Measures on Goods Trade in the East African Community.� Africa Trade Policy Note 7, World Bank, Washington, DC. 237 Portugal-Perez, Alberto, José-Daniel Reyes, and John S. Wilson. 2010. “Beyond the Information Technology Agreement: Harmonisation of Standards and Trade in Electronics.� The World Economy 33(12): 1870–97. 107 134. The crisis raised fears of widespread protectionism but researchers demonstrated that this was a dog that did not bark. Only a handful of countries, such as Argentina, China, Malawi, Russia, and Turkey imposed higher tariffs on products with significant trade flows. But the crisis did accentuate a longer-term trend of increasing recourse to antidumping, safeguard, and countervailing duty policies among some countries — as revealed by the World Bank’s new Database on Temporary Barriers to Trade. 238 Countries sometimes used antidumping and safeguard exceptions to unwind commitments to lower tariffs in the face of domestic political- economic pressure. 239 Moreover, these barriers are increasingly a “South-South� phenomenon, with China, in particular, a victim of actions by other developing countries. 240 Still, the rise in tariffs and antidumping duties is estimated to have jointly caused trade to drop only by about US$43 billion, which is less than 2 percent of the decline in global trade. 241 Aggregating across all product categories, the evidence suggests that the collapse in trade was caused primarily by a synchronized demand-side shock, but supply-side frictions did play a role within manufacturing. 242 135. Who gets the protection? It is well-known that declining industries are more likely to receive trade protection, yet standard economic approaches suggest that protection should be given to expanding sectors. To explain this anomaly, a new study introduced concepts of loss aversion from behavioral economics to better explain actual policy choices. A clear prediction emerges: when an industry first begins to decline, significant protection is provided to isolate it from reductions in the world price. However, if the industry continues to decline, policy approaches free trade. 243 New firm-level data is yielding insights on how firms begin to export and what sustains exports… 136. Understanding what drives and sustains exports success has proved elusive, but a new dataset being constructed by the trade research group is helping. For example, firm-level data on the nontraditional agriculture sector in Peru, which grew seven-fold from 1994 to 2007, reveals new products are typically introduced by large experienced exporters. These introductions 238 Bown, Chad P., ed. 2011. The Great Recession and Import Protection: The Role of Temporary Trade Barriers. London, UK: CEPR and Washington, DC: World Bank. 239 Bown, Chad P., and Patricia Tovar. 2011. “Trade Liberalization, Antidumping, and Safeguards: Evidence from India’s Tariff Reform.� Journal of Development Economics 96(1): 115–25. 240 Bown, Chad P. 2011. “Taking Stock of Antidumping, Safeguards and Countervailing Duties, 1990–2009.� The World Economy 34(12): 1955–98. (Lead article). 241 Kee, Hiau Looi, Cristina Neagu, and Alessandro Nicita. Forthcoming. “Is Protectionism on the Rise? Assessing National Trade Policies during the Crisis of 2008.� Review of Economics and Statistics. (Based on Policy Research Working Paper 5274, World Bank, Washington, DC.) 242 Haddad, Mona, Ann Harrison, and Catherine Hausman. 2010. “Decomposing the Great Trade Collapse: Products, Prices, and Quantities in the 2008–2009 Crisis.� NBER Working Paper 16253, National Bureau of Economic Research, Cambridge, Mass. 243 Freund, Caroline, and Çağlar Özden. 2008. “Trade Policy and Loss Aversion.� American Economic Review 98(4): 1675–91. 108 provoke tremendous firm entry and exit in the export sector. 244 Similarly, data for Costa Rica for the period 1997–2007 reveal that on average, about 30 percent of firms in each year tend to exit export activities, and a similar percentage of firms enter. Even though exiting and entering firms tend to be smaller than incumbent firms in terms of export value, in the long run new product- firm combinations (i.e., product-firm combinations not present in 1997) account for almost 60 percent of the value of exports in 2007.245 A study of Mexican trade integration under NAFTA also shows intense product churning within firms, and that new exporters enter foreign markets with a small export volume and a small number of varieties, most of which were previously sold at home. 246 137. Africa is no different. A study of Malawi, Mali, Senegal and Tanzania reveals a high degree of experimentation with new products and markets associated with low survival rates, as in high- and middle-income countries. Survival probability rises with the number of firms exporting the same product to the same destination from the same country, pointing to the existence of cross-firm synergies, which may be driven by information spillovers. Firms that are more diversified, in terms of products and in terms of markets, are more likely to be successful and survive beyond the first year. 247 …and on the relationship between trade and economic performance 138. It is now well-established that exporters tend to be more productive than domestic firms. However, these premiums are found to vary considerably across countries, being higher in countries with lower export participation rates, with more restrictive trade policies, and in countries exporting to relatively more distant markets. 248 139. A study finds only weak evidence that “what you export matters� and suggests that “how� any particular good is produced merits more attention from a policy perspective. For example, natural resources are not necessarily a curse, and some countries have leveraged them into sustainable and diversified development while others have not. Different goods do appear to have potential for growth in export quality measured by unit value, but poor countries show lower growth even conditioning on goods composition. “High tech� goods such as computers 244 Freund, Caroline, and Martha Denisse Pierola. 2010. “Export Entrepreneurs: Evidence from Peru.� Policy Research Working Paper 5407, World Bank, Washington, DC. 245 Lederman, Daniel, Andrés Rodríguez-Clare, and Daniel Yi Xu. 2011. “Entrepreneurship and the Extensive Margin in Export Growth: A Microeconomic Accounting of Costa Rica’s Export Growth during 1997–2007.� World Bank Economic Review 25(3): 543–61. 246 Iacovone, Leonardo, and Beata S. Javorcik. 2010. “Multi-Product Exporters: Product Churning, Uncertainty and Export Discoveries.� Economic Journal 120(544): 481–99. 247 Cadot, Oliver, Leonardo Iacovone, Martha Denisse Pierola, and Ferdinand Rauch. 2011. “Success and Failure of African Exporters.� Policy Research Working Paper 5657, World Bank, Washington, DC. 248 Fernandes, Ana M. 2008. “Understanding Cross-Country Differences in Exporter Premia: Comparable Evidence for 14 Countries.� Review of World Economics 144(4): 596–635. 109 can be the product of highly skilled human capital, but can also involve rote assembly work with little dynamic growth potential. 249 140. Increased competition from low-price producers in China and India has induced manufacturing firms in other emerging economies to position themselves in domestic and international markets by offering upgraded and differentiated rather than generic labor-intensive products. For example, a rich dataset of Chilean manufacturing plants reveals a positive and robust effect of import competition on product quality. 250 141. The link between market incentives and technology adoption is more nuanced. Survey data for 7,000 firms in 28 countries in Eastern Europe and Central Asia reveal that while stronger consumer pressure is significantly associated with technology adoption, competitor pressure is not, suggesting that it is primarily firms with rents that are able to adopt new technologies. Foreign-owned firms exhibit significantly better technology adoption outcomes, but privatized firms with domestic owners do not. 251 142. The impact of international integration on the labor market is a central concern for policy makers. New research suggests that the impact of exports depends on their destination. One reason is that exporting to high-income countries requires quality upgrades that are skill- intensive. A study of Argentine manufacturing firms during 1998–2000 reveals that firms induced by the Brazilian currency devaluation of 1999 to shift from Brazilian to high-income markets hired a higher proportion of skilled workers and paid higher average wages than other exporters (to non-high-income countries) and domestic firms. 252 Trade policy is taking newer pro-active forms and these efforts need to be informed by better evidence 143. In developing countries and the World Bank’s work, there is a shift away from economy- wide reforms of tariffs and trade restrictions toward focused interventions to facilitate trade and promote exports. But there is limited evidence on whether such interventions work and why. A new book on the impact evaluation of trade-related policy interventions took the first steps toward more rigorous evaluation of trade-related interventions, drawing upon experience in other areas (labor, education, health). Systematically building impact evaluation into trade projects could lead to better policy design and a more credible case for “aid-for-trade.� 253 249 Lederman, Daniel, and William F. Maloney. Forthcoming. Does What You Export Matter? In Search of Empirical Guidance for Industrial Policies. Latin American Development Forum. Washington, DC: World Bank. 250 Fernandes, Ana M., and Caroline Paunov. 2009. “Does Tougher Import Competition Foster Product Quality Upgrading?� Policy Research Working Paper 4894, World Bank, Washington, DC. 251 Correa, Paulo G., Ana M. Fernandes, and Chris J. Uregian. 2010. “Technology Adoption and the Investment Climate: Firm-Level Evidence for Eastern Europe and Central Asia.� World Bank Economic Review 24(1): 121–47. 252 Brambilla, Irene, Daniel Lederman, and Guido Porto. Forthcoming. “Exports, Exports Destinations, and Skills.� American Economic Review. 253 Cadot, Olivier, Ana M. Fernandes, Julien Gourdon, and Aaditya Mattoo, eds. 2011. Where to Spend the Next Million? Applying Impact Evaluation to Trade Assistance. London: CEPR and Washington, DC: World Bank. 110 Trade facilitation efforts must target infrastructure and delay 144. As tariffs have declined, it has become evident that protection is not the only impediment to trade. Weakness in infrastructure—both hard (e.g., ports and roads) and “soft� (e.g., customs and other regulations)—also hurt trade. Trade costs are higher in Africa than in other regions, and trade facilitation could have a powerful effect on trade. 254 Similarly, improving the quality of physical infrastructure so that Egypt’s indicator increases half-way to the level of Tunisia would increase its exports by more than 10 percent, roughly equal in impact to a 7.5 percent cut in tariffs faced by Egyptian exporters in destination markets. 255 A 1-percent increase in aid-for- trade facilitation (about US$220 million in 2008) could potentially result in US$290 million of additional exports. 256 145. The Bank has developed new measures of trade costs, such as the Logistics Performance Index (LPI), which are spurring efforts to facilitate trade through reforms that encompass behind- the-border measures. 257 Estimates suggest that an increase in the LPI of low-income countries to the middle-income country average would increase imports by 8.5 percent and exports by 15 percent. 258 Apart from explicit trade costs, “trade delayed is trade lost.� Using newly collected data on the days it takes to move standard cargo from the factory gate to the ship in 98 countries, it was estimated that each additional day that a product is delayed prior to being shipped reduces trade by more than 1 percent. 259 Well-designed export promotion can help 146. The number of national export promotion agencies (EPAs) has tripled over the past two decades. Many EPAs were retooled, partly in response to critiques of their efficacy in developing countries. New survey data covering 103 developing and developed countries, suggest that today’s EPAs and their strategies have a statistically significant effect on exports. EPA services help overcome foreign trade barriers and address information problems associated with exports 254 Portugal Perez, Alberto, and John S. Wilson. 2009. “Trade Facilitation in Africa: Why Reform Matters?� World Trade Review 8(3): 379–416. 255 Portugal Perez, Alberto, and John S. Wilson. Forthcoming. “Export Performance and Trade Facilitation Reform: Hard and Soft Infrastructure.� World Development. (Based on Policy Research Working Paper 5261, World Bank, Washington, DC.) 256 Helble, Matthias, Catherine L. Mann, and John S. Wilson. Forthcoming. “Aid-for-Trade Facilitation.� Review of World Economics. (Based on Policy Research Working Paper 5064, World Bank, Washington, DC.) 257 Arvis, Jean-François, Monica Alina Mustra, Lauri Ojala, Ben Shepherd, and Daniel Saslavsky. 2010. Connecting to Compete: Trade Logistics in the Global Economy: The Logistics Performance Index and its Indicators. Washington, DC: World Bank. 258 Hoekman, Bernard, and Alessandro Nicita. 2011. “Trade Policy, Trade Costs and Developing Country Trade.� World Development 39(12): 2069–79. 259 Djankov, Simeon, Caroline Freund, and Cong S. Pham. 2010. “Trading on Time.� Review of Economics and Statistics 92(1): 166–73. 111 of heterogeneous goods. There are also strong diminishing returns, suggesting that as far as EPAs are concerned, small is beautiful. 260 147. A new book examines the sudden and abrupt drop in international trade and the reported trade finance “gap� during the 2008–09 financial crisis. It finds that trade finance was not the main driver behind the 2008 trade collapse, but that small and medium enterprises were particularly vulnerable to the tightening of trade finance conditions. The swift response of the international community in maintaining specific programs to support vulnerable segments helped. 261 148. Exchange rate policy also matters. Export surges in developing countries tend to be preceded by a large real depreciation, which leaves the exchange rate significantly undervalued. Depreciation is associated with a significant reallocation of resources in the export sector, and stimulates entry into new export products and new markets. These new exports are important, accounting for over 40 percent of export growth, on average, during the surge in developing countries. 262 149. The impact of political instability and economic mismanagement on trade performance cannot be easily reversed, as a study of Zimbabwe shows. Over the last decade, traditional surpluses in agricultural products, industrial raw materials, and tourism have either diminished or disappeared turning Zimbabwe into a net importer of agricultural products. Macroeconomic stability was restored in 2009 after more than a decade of inflation and hyperinflation. But it is taking time for economic reform to revive even the exports that are in line with Zimbabwe’s endowments of natural resources, human capital, and natural environment. 263 Services reform may also be an effective form of industrial policy 150. There is increasing evidence that services liberalization is a major potential source of gains in economic performance. 264 Firm-level data from the Czech Republic, show a positive relationship between services sector reform and the performance of domestic firms in downstream manufacturing sectors. 265 Allowing foreign entry into services industries appears to be the key channel, and foreign acquisitions of Czech services providers result in profound 260 Lederman, Daniel, Marcelo Olarreaga, and Lucy Payton. 2010. “Export Promotion Agencies: Do They Work?� Journal of Development Economics 91: 257–65. 261 Chauffour, Jean-Pierre, and Mariem Malouche, eds. 2011. Trade Finance during the Great Trade Collapse. Washington, DC: World Bank. 262 Freund, Caroline, and Martha Denisse Pierola. 2012. “Export Surges.� Journal of Development Economics 97(1): 387–95. 263 Kaminski, Bartlomiej, and Francis Ng. Forthcoming. “Impact of Zimbabwe’s Decade of Dynamics in Reverse on its Exports Performance.� African Development Review. 264 Francois, Joseph, and Bernard Hoekman. 2010. “Services Trade and Policy.� Journal of Economic Literature 48(3): 642–92. 265 Arnold, Jens M., Beata S. Javorcik, and Aaditya Mattoo. 2011. “Does Services Liberalization Benefit Manufacturing Firms? Evidence from the Czech Republic.� Journal of International Economics 85(1): 136–46. 112 changes in the labor productivity and sales of acquired firms. Similarly, a study for Chile shows that forward linkages from foreign direct investment (FDI) in services explain 7 percent of the observed increase in manufacturing users’ total factor productivity. 266 FDI is shown to foster innovation activities in manufacturing and offers opportunities for laggard firms to catch up with industry leaders. It has been estimated that a large part of the gains for Russia from WTO accession will be due to the liberalization of barriers against multinational service providers.267 As most barriers to foreign investment today are not in goods but in services sectors, the findings of these studies may strengthen the argument for additional reform in this area. New data are improving our understanding of the patterns of migration and its impact on development 151. A new database of bilateral migrant stocks spanning 1960–2000, and drawing upon data from over 1,000 national censuses and population registers, presents a comprehensive picture of bilateral global migration. The global migrant stock increased from 92 million in 1960 to 165 million in 2000. Migration between developing countries dominates but migration from developing to developed countries is growing faster. The United States is the most important migrant destination in the world, home to one-fifth of the world’s migrants and the top destination for migrants from some 60 sending countries. Migration to Western Europe has come largely from elsewhere in Europe. The oil-rich Persian Gulf countries emerge as important destinations for migrants from the Middle East and North Africa and South and Southeast Asia. 268 Migration flows are shaped by a combination of self-selection and out-selection mechanisms and existing diasporas are among the most important determinant of migration patterns. 269 152. In terms of the development impact of migration, brain drain has long been a common concern for migrant-sending countries, particularly for small countries where high-skilled emigration rates are highest. Detailed surveys were conducted to track academic high-achievers from five countries to wherever they moved. The results show that there are high levels of emigration and of return migration among the highly skilled and the income gains and remittances are large. 270 266 Fernandes, Ana M., and Caroline Paunov. 2012. “Foreign Direct Investment in Services and Manufacturing Productivity: Evidence for Chile.� Journal of Development Economics 97(1): 305–21. 267 Rutherford, Thomas F., and David G. Tarr. 2010. “Regional Impacts of Liberalization of Barriers against Foreign Direct Investment in Services: The Case of Russia’s Accession to the WTO.� Review of International Economics 18(1): 30–46. 268 Özden, Çağlar, Chris Parsons, Maurice Schiff, and Terrie Walmsley. 2011. “Where on Earth is Everybody? The Evolution of Global Bilateral Migration 1960–2000.� World Bank Economic Review 25(1): 12–56. 269 Beine, Miche, Frederic Docquier, and Çağlar Özden. 2011. “Diasporas.� Journal of Development Economics 95(1): 30–41. 270 Gibson, John, and David McKenzie. Forthcoming. “The Economic Consequences of ‘Brain Drain’ of the Best and Brightest: Microeconomic Evidence from Five Countries.� Economic Journal. (Based on Policy Research Working Paper 5394, World Bank, Washington, DC.) 113 153. Striking differences exist among highly educated immigrants in the United States, even after controlling for age, experience, and level of education. Educated immigrants from Latin American and Eastern Europe are more likely to end up in unskilled jobs than those from Asia and industrial countries. “Under-placed� migrants suffer primarily from low (or poorly transferable) skills and many problems might be reduced with better sharing of information on labor market conditions and recognition of workers’ qualifications. 271 154. Seasonal migration programs are widely used around the world, and are increasingly seen as offering a potential “triple-win�—benefiting the migrant, sending country, and receiving country. New Zealand’s Recognized Seasonal Employer program was launched in 2007 with an explicit focus on development in the Pacific alongside the aim of benefiting employers at home. A multi-year prospective evaluation measured the impact of participation in this program on households and communities in Tonga and Vanuatu and found positive development impacts. It has increased income and consumption of households, allowed purchase of more durable goods, and increased the subjective standard of living and child schooling in Tonga. 272 155. The importance of ethnic networks for international trade has been recognized, but their impact on foreign direct investment (FDI) has not. The presence of migrants can stimulate FDI by promoting information flows across international borders and by serving as a contract enforcement mechanism. New research reveals that U.S. FDI abroad is positively correlated with the presence of migrants from the host country, especially those with tertiary education. 273 Another study on Albania also finds that past household migration experience exerts a positive impact on the probability of owning a non-farm business; however, while one additional year in Greece increases the probability of household business ownership by roughly 6 percent, a similar experience in Italy or farther destinations raises the probability by over 25 percent. 274 156. Diaspora bonds can be a powerful financial instrument for mobilizing savings of emigrants to finance specific public and private sector projects, as well as to help improve the debt profile of their home countries. There are risks associated with debt denominated in foreign currency and consideration must be given to prudential risk management before taking on additional debt. Building on extensive analytical work, pilot programs are proposed for funding infrastructure, education and community development projects via diaspora bonds. 275 271 Mattoo, Aaditya, Ileana Cristina Constantinescu, and Çağlar Özden. 2008. “Brain Waste? Educated Immigrants in the US Labor Market.� Journal of Development Economics 87(2): 255–69. 272 Gibson, John, and David McKenzie. 2011. “The Development Impact of a Best Practice Seasonal Worker Policy.� Policy Research Working Paper 5488, World Bank, Washington, DC. 273 Özden, Çağlar, Beata S. Javorcik, Cristina Ileana Neagu, and Mariana Spatareanu. 2010. “Migrant Networks and Foreign Direct Investment.� Journal of Development Economics 94(2): 231–41. 274 Kilic, Talip, Calogero Carletto, Benjamin Davis, and Alberto Zezza. 2009. “Investing Back Home: Return Migration and Business Ownership in Albania.� Economics of Transition 17(3): 587–623. 275 Ketkar, Suhas, and Dilip Ratha, eds. 2009. Innovative Financing for Development. Washington, DC: World Bank. 114 157. Immigrants in Rome or Paris are more visible to the public eye than the Italian or French engineers in Silicon Valley. Nevertheless, public fears that immigration worsens income distribution may be misplaced, especially in European countries. A new dataset on migration flows by education levels for the period 1990–2000, reveals that both emigrants and immigrants are more skilled than non-migrants. Therefore, immigration generally improved the income distribution of European countries while emigration worsened it by increasing the wage gap between the high- and low-skilled natives. 276 Whither international cooperation? 158. The Doha Development Agenda (DDA) is in limbo. A new book helps a Doha-weary world make informed choices as it faces a difficult “trilemma�: to implement all or part of the draft agreements as they stand today; to modify them substantially; or to dump Doha and start afresh. Current Doha proposals—even after allowing for flexibilities such as for sensitive and special products—would cut applied tariffs on agricultural and non-agricultural (NAMA) goods by around 20 percent; in agriculture, abolish export subsidies and sharply reduce maximum levels of domestic support, especially in the European Union and the United States: and by cutting bound tariffs (an average of 27 percent in agricultural and 46 percent in non-agricultural goods) reduce the scope for future protection. 277 159. The difficulties involved in completing the Doha Development Agenda raise important questions about whether fundamental reform of the WTO system is required, both in terms of how the traditional agenda is negotiated and what new issues need to be addressed. 278 For example, less ambitious tariff-cutting formulae in goods may generate less pressure for exceptions; addressing export restrictions in agriculture may make import liberalization easier; and greater regulatory cooperation in services may facilitate market access negotiations. Furthermore, a negotiating agenda that speaks to 21st Century concerns could include issues such as food and energy security, currency undervaluation and climate change-related trade issues, and involve greater cooperation between international organizations. 160. Even as multilateral negotiations are stalled, new regional initiatives are emerging. Some are concerned that Free Trade Areas (FTAs) reduce the trade opportunities of non-members. A study of 10 Latin American countries finds that preferential tariff reduction actually leads to reductions in all external tariffs. External liberalization is greater if preferential access is granted to important suppliers. These “complementarity effects� of preferential liberalization do not arise in customs unions, suggesting an important reason to prefer FTAs over Customs Unions. 279 276 Docquier, Frédéric, Çağlar Özden, and Giovanni Peri. 2010. “The Wage Effects of Immigration and Emigration.� NBER Working Paper 16646, National Bureau of Economic Research, Cambridge, Mass. 277 Mattoo, Aaditya, and Will Martin. 2011. Unfinished Business? The WTO’s Doha Agenda. London: CEPR and Washington, DC: World Bank. 278 Mattoo, Aaditya, and Arvind Subrmanian. 2009. “From Doha to the Next Bretton Woods.� Foreign Affairs 88(1): 15–26. 279 Freund, Caroline, Estevedordael Antoni, and Emanuel Ornelas. 2008. “Does Regionalism Promote External Trade Liberalization Toward Nonmembers?� Quarterly Journal of Economics 123(4): 1531–75. 115 161. While Latin American countries have primarily used formal regional trade treaties as the main channel of integration, East Asian countries have been “integrating via markets� before negotiating formal agreements. One interpretation of the relative effectiveness of the East Asian approach is that regional trade agreements often serve multiple constituents, and integrating via markets empowers the outward-looking economic interests. 280 162. As climate change negotiations enter a post-Durban phase, new research examines the trade consequences of mitigation actions. While manufacturing output and exports in low-carbon intensity countries such as Brazil are unlikely to be adversely affected, in high-carbon intensity countries, such as China and India, even a modest agreement could significantly depress manufacturing output and exports. 281 If industrial countries impose border taxes on imports from countries with lower carbon prices, their impact would depend on their design. A tariff based on the carbon content of imports would have serious consequences for trading partners; for example, China’s manufacturing exports would decline by one-fifth and those of all low- and middle-income countries by 8 percent. In contrast, a tariff based on the carbon content in domestic production would broadly address the competitiveness concerns of producers in high- income countries and less seriously damage developing country trade. 282 Prepared by Aaditya Mattoo, Research Manager, Development Research Group, Development Economics Vice Presidency. 280 Aminian, Nathalie, K. C. Fung, and Francis Ng. 2009. “A Comparative Analysis of Trade and Economic Integration in East Asia and Latin America.� Economic Change and Restructuring 42(1): 105–37. 281 Mattoo, Aaditya, Arvind Subramanian, Dominique van der Mensbrugghe, and Jianwu He. Forthcoming. “Can Global De-Carbonization Inhibit Developing Country Industrialization?� World Bank Economic Review. (Based on Policy Research Working Paper 5121, World Bank, Washington, DC.) 282 Mattoo, Aaditya, Arvind Subramanian, Dominique van der Mensbrugghe, and Jianwu He. 2009. “Reconciling Climate Change and Trade Policy.� Policy Research Working Paper 5123, World Bank, Washington, DC. 116 ANNEX B. DEC VPU BUDGET 1. The Development Economics (DEC) VPU budget allocation remained essentially flat in real terms at around $52 million for the period FY 2009–2011 (Table B.1). DEC’s share of the total World Bank net administrative budget during the three fiscal years remained steady at 2.9 percent, but is expected to decline slightly in FY 2012 to 2.8 percent. The allocated budget for the Research Group alone accounts for 1.2 percent of the total Bank net administrative budget. Table B.1: Overview of DEC VPU Budget, FY 2009–2012 Budget Allocation by Fiscal Year (in constant 2012 dollars) DEC VPU Budget FY12 FY09 FY10 FY11 Projection (Millions) DEC allocated Bank budget1/ 52.5 52.3 52.9 51.6 Percentage of total Bank net admin. budget 2.9% 2.9% 2.9% 2.8% Adjustments2/ 9.9 12.3 7.9 5.0 DEC total bank resources 62.4 64.5 60.9 55.7 Of which: DEC Research Group–allocated Bank budget 22.3 21.6 20.4 20.4 Percentage of total Bank Net admin. budget 1.2% 1.2% 1.1% 1.1% Note: DEC-managed Bank executed trust funds3/ 13.9 20.3 27.7 24.1 Bank net administrative budget 1,787.1 1,788.6 1,810.2 1,823.3 Reflects the BB portion of the program budget per the FY 2009–2012 budget documents, including the research support budget in 1/ DEC. 2/ Adjustments include budget transfers that are executed during the fiscal year based on work program agreements, inter-VPU cross- support, reimbursable budget and fee income earned, contingency allocations, and any authorized overruns. 3/ Reflects trust fund disbursements. Source: Business Warehouse 2. Given the 2 percent real decline in budget allocation from FY 2009 to FY 2012 (or the 5 percent real decline since FY 2007, as shown in Figure B.1), DEC has turned more often to external resources, with the Knowledge for Change Program instrumental in supporting more innovative research activities that would otherwise remain unfunded. The volume of DEC trust 117 fund disbursements has grown by some 23 percent on average over the three years, having peaked in FY 2011. Heavy reliance on trust funds to support Bank research, however, is neither desirable nor sustainable for carrying out a robust and multi-year research program. Figure B.1—Trend in DEC’s Allocated Bank Budget and Trust Funds, FY 2007–2013 The Bank is under-investing in research and data compared to some other organizations 3. Between FY 2007 and FY 2012, the DEC VPU budget declined by 5 percent in real terms. This decline in allocated Bank budget has required significant tradeoffs, with direct implications for research. Budget pressures have necessitated greater selectivity among research priorities, contributing to some deterioration over recent years of research capacity in some core areas. Since researchers’ involvement in operational work improves familiarity with research and its use, this is likely to have adversely affected perceptions of research in parts of Bank operations. A key challenge for research managers in staffing their teams is to attempt to anticipate the future research needs of clients, while continuing to deliver on what are often multi-year research activities, amid demands for more flexible staffing in the face of budget constraints. Greater emphasis on producing more data and tools in recent years have also put significant pressure on the budget, since these wholesale research activities are not only resource intensive themselves, but require strong in-house research in all core areas. Examples of how researchers produce new datasets and tools as an integral part of their research are given in Chapter 3 above. 118 4. Benchmarking the Bank’s allocation to research and data activities compared to that of other large institutions reveals that the Bank has been significantly under-investing in these areas. For example, the IMF’s research and data budget is 6.8 percent of its administrative budget, and the United States Federal Reserve Board of Governors spends about 12 percent of its operations budget on research and statistics (Figure B.2 and Table B.2). 5. While valid comparators beyond similar institutions are difficult to identify, it is interesting to note that the Bank also spends much less on research than the private sector on average spends on research and development activities. For example, OECD data show that in 2008 (the latest year available), manufacturing firms in the United States spent on average 10.5 percent of value added on R&D, while in knowledge-intensive industries such as information technology the share has been as high as 28 percent. Another comparison comes from U.S. federal funding of academic research, which grew by 2.8 percent per year in real terms during the 2000s, and even increased by 22 percent cumulatively between 2006 and 2009, despite tough economic times. By these standards, the Bank is substantially under-investing in research and data creation. Figure B.2—Share of Budget Allocated to Research and Data 0.14 Research Budget as a % of Total Budget* 0.12 12.1% 0.1 0.08 0.06 6.8% 0.04 0.02 2.9% 0 The World Bank International Monetary Fund Board of Governors Federal Reserve, United States * Total Buget: World Bank(2011) and IMF(2010) - Administrative Budget; US Federal Reserve(2009) - Board Operations Budget 119 Table B.2: Share of Research and Data Budget in Different Organizations Research and Data Name of the Institution1 Share of What Budget? Budget Share The World Bank 2.9% Total Bank net administrative budget (2011) International Monetary Fund 6.8% Total administrative budget (2010) Board of Governors Federal 12.1% Total Board operations budget (2009) Reserve, United States 1. Research budget for international organizations are obtained from various sources including Business Warehouse (World Bank) and respective annual reports. 120